main
100
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Mike Kaganski
|
a36c7459d1 |
Fix unit test testPieTextArea
CppunitTest_oox_shape failed in the gerrit_linux_gcc_main and
gerrit_linux_clang_dbgutil_main builds like
oox/qa/unit/shape.cxx:1170:testPieTextArea::TestBody
double equality assertion failed
- Expected: 6673
- Actual : 6691
- Delta : 10
The test from 2ac8858b10699c897dd530c8b1434bd9604a4aff (tdf#149918 fix
pie in presetShapeDefinitions.xml, 2026-10-06) checked the BoundRect of
the shape, which includes the ink of its text; text layout changes merged
meanwhile probably moved that ink. Check the text frame that the preset
defines instead; it does not depend on fonts or line metrics.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: Ica16e4457cc150e9115dfc9994abd3f801778621
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/14447
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
(cherry picked from commit 0bcc06cdc592bf7c9828b694b00acb7fed07eccf)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211889
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
4f8889f928 |
vcl: anchor the object-stream test on /Type/ObjStm
testEncryptedTaggedUsesObjectStreams searched the whole exported file for /ObjStm, which binary data can hold, and every later search starts from that match. The writer emits /Type/ObjStm, with /Length and the stream data after it in the same dictionary. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ibb1417ee9bd399ea183883d2849f02a4e3d945c4 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13491 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 8171259354f2e36fddcdb815d393d7d60c303e01) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211702 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
c5592ec7db |
tdf#163240 sw: build a link's rectangles from its line pieces
Exporting a hyperlink laid out down the page, over more than one line,
produced link annotations a PDF/UA validator rejects: on a vertical page
whose link wraps, two of the three annotations carried no structure
element, and veraPDF reported both under ISO 14289-1 7.18.5:
A Link annotation is an Artifact or is nested within null tag
(standard type = null) instead of Link
Export a document whose hyperlink ends in a ruby and runs over several
lines down the page to PDF with tagged PDF enabled, then run veraPDF
over the result in PDF/UA-1 mode to see it.
An annotation is tied to a structure element by the text painted inside
it, so one holding no text of any line is tied to nothing. That is what
the region CalcFrameRects returns can hand out: it covers the selection
in whatever rectangles fall out of inverting the pieces it was built
from, and one of those can span several lines while another holds a ruby
and none of the text beside it.
Let CalcFrameRects return those pieces as well, and the holes it cuts
afterwards, and take the shape from the pieces and the area from the
region: the pieces follow the lines and, once the overlapping ones are
joined, no longer overlap each other, so every part of the region falls
in exactly one of them. What tdf#157816 requires still decides each
part, and the parts a piece has left are joined unless the join would
reach over a hole - a fly, a drop cap - that the selection excludes on
purpose.
Every annotation now holds text of the lines it covers, so every one of
them is tagged, and a ruby sits inside the rectangle of the line that
carries it. The narrowing to a portion's band, the growing over a ruby
portion and the joining of tiling rectangles that
Ide896d00c7619b772802762e85abfb6e2cf7ef0a ("tdf#163240 sw: let a link's
rectangle cover the ruby it ends in") added are all gone. On the
horizontal document that change was measured against, the exported
rectangles are the same to the last digit.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: I9faf2bc10af491576cc2f4ba5f3e3433eb03d16c
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13767
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
(cherry picked from commit e77b1d25d3601315f80169e4e811d74ef2ff0374)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211704
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
||
|
Mike Kaganski
|
8c112eb826 |
tdf#163240 sw: let a link's rectangle cover the ruby it ends in
A ruby makes its line taller than the line's own text, so for a
hyperlink that ends inside one the selection region for that line was
split into two rectangles: the band holding the line's text, which runs
on under the ruby's base, and above it a band reaching only as far as
the ruby begins. Neither covered the phonetic annotation standing above
the base. Each became a link annotation in the exported PDF, so clicking
the phonetic annotation did nothing, while a 6.9pt annotation stood over
blank space carrying no structure element, which veraPDF reported under
ISO 14289-1 7.18.5:
A Link annotation is an Artifact or is nested within null tag
(standard type = null) instead of Link
Export a document whose hyperlink ends in a ruby to PDF with tagged PDF
enabled, then run veraPDF over the result in PDF/UA-1 mode to see it.
The same document with its text running down the page draws the clause
twice.
The rectangle a hit test returns for a character is as tall as the whole
line, ruby and all, so GetCursorRectsContainingText, which
|
||
|
Mike Kaganski
|
4998ea8631 |
sw: emit a Ruby's RB before its RT
ISO 32000-2 Table 369 makes the order normative: a Ruby shall contain one RB followed by an RT, or by RP, RT and RP. A ruby's two sub-lines paint in visual order, though, and each opened its structure element as it painted, so an exported ruby held its RT before its RB. PAC 2026 reports that as two errors per ruby; veraPDF does not report it. An element is ordered among its siblings where its type is set, not where it is painted, so both kids are created and typed before either sub-line paints, and each sub-line then enters the kid already created for it. Which sub-line carries the annotation follows OnTop, while bRubyTop only reflects where a squared grid puts the annotation and is true everywhere else. Keying the role off bRubyTop tagged the base of a ruby positioned below outside a squared grid, or to the right, as the RT and its annotation as the RB, so both the role and the element follow OnTop now. A squared grid with the ruby text below was already right, bRubyTop and OnTop agreeing there. The kids are created only where BeginInlineStructureElements would enter them: a ruby in the repeated headline row of a follow table would otherwise leave an empty RB and RT on the table, once per following page. testLinkWithRuby asserted only that both kids are present, the order being wrong when it was written; it compares them in order now. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I8e38a3544008f79939d7d0795c4808a1a72caad7 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13402 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 65c62bc31aff793465f31ab6172c118921fe9d51) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211690 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
eec52071ef |
tdf#157028 vcl: use object streams in an encrypted file too
Object streams and a cross-reference stream shrink a tagged PDF, but
I1e27dc3a8f43aa8b6c75ae930615c8d4b4645627 ("tdf#157028 vcl: write the
structure tree into object streams") left an encrypted export out of
both, so a password kept the whole structure tree uncompressed. The
spreadsheet reported in tdf#166233, attachment 171182, exported tagged
at 2368642 bytes, and at 13465120 with a password. The protected export
is 2389620 bytes now, and what it has over the plain tagged one is what
encryption itself costs.
This writer encrypts a string as it builds the object. ISO 32000-2 7.5.7
forbids encrypting a string in an object stream separately: the stream
that carries it is encrypted instead. So the strings of a stream-bound
object go in plain now, and the object stream is encrypted as a whole,
with the cipher set up for the stream's own object number rather than
for the objects inside. The stream's /Length counts what the cipher
adds. Only the structure elements and the ID tree put strings in a
stream, so only those two builders ask getStringEncryptor for one, and
it returns nothing while the file uses object streams.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: Ia6a063244baf456d7c7368a6bba7a8cbb96e31b0
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13370
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
(cherry picked from commit 59193295323917bd065ea3a235f7653ad7f9b04b)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211655
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
24f40c4b11 |
tdf#163272 sw,drawinglayer,vcl: give anchored content its own /Lang
With Tagged PDF on, a paragraph in a language other than the document's gets /Lang on its structure element. What is anchored in the paragraph is tagged inside that same element and inherits the attribute, so a frame's text and its description would be declared the paragraph's language. 0d2d9e656d87647f904f1a463ed9b879c2bfb10c (tdf#163272 sw: give a paragraph its own /Lang, 2026-09-21) avoided that by putting no /Lang on such a paragraph, on a check of one node of a text frame that can cover several. Its text still carried the language, on a Span, while what was anchored in it specified none and inherited the document's. In a German document that declared a shape's English text German. See https://gerrit.collaboraoffice.com/c/online/+/13117/comment/89a7866d_907a2957/ The paragraph specifies its own language now, and a frame anchored in it no longer inherits it. A Figure, a Formula and a frame's Division specify the document's where the paragraph's differs, which is what the frame's own paragraphs are measured against. A drawing shape's text is tagged by drawinglayer, which takes the locale of the drawn portions and specifies it where they agree; such a shape is nested in its anchor paragraph in Writer and in no such paragraph elsewhere, so there is nothing to compare against. A drawing shape's own element specifies the document's language too. svx anchors it under the anchor paragraph, drawinglayer opens it there, and it carries the shape's title and description. Without the attribute, a described shape in a paragraph that specifies one would have its description declared the paragraph's language, and its text likewise where the portions disagree. testParagraphLanguage compared a structure type against a language and quoted no text, so nothing identified the words affected. It lists every element holding text or a description now, against the language it specifies, an entry reading "P de-DE: Ein deutscher Absatz mit einem Bild."; one specifying none is listed without a language, so correct inheritance is asserted too. PDFiumPageObject gained getMarkedContentID to join a page object's text to its element. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I1eecf0b56f09fc6d144894910a9037d89884dd3b Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13239 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 898a2e1b3a6fc07d7b43315b2f4781b32f1a20d3) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211609 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
d00d192770 |
tdf#157028 vcl: write the structure tree into object streams
Turning Tagged PDF on added 2.4 MB to the reporter's document, 2.0 MB of that its structure tree: 12562 elements, each its own top-level object, uncompressed, and each costing a further 20 bytes in the cross-reference table. The elements go into /Type/ObjStm streams of 200 now, with the parent and ID trees. The cross-reference has to become a /Type/XRef stream with them, because a table cannot address an object inside another object. PDF 1.5 introduced both, so a PDF/A-1b file keeps the plain form. So does an encrypted file, since a string in an object stream shall not be separately encrypted, and this writer encrypts strings as it builds the object. So does a hybrid file, because sdext reads the embedded original document out of a trailer. The reporter's document exported 3157118 bytes tagged before the change and 1086230 after, against 769448 untagged either way. An 8300-row Writer table exported 4017140 bytes tagged before and 500967 after, against 73183 untagged. testTdf159066 and testTdf173162 grep the exported bytes for something the structure tree carries, /Alt< and the /Namespaces array. They inflate the object streams first now. testListLabelPDFExport counted its List element in the token stream, and asks GetObjects() for it. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I1e27dc3a8f43aa8b6c75ae930615c8d4b4645627 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13209 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit ad4bb8ae080aba98080dfaabd688dfdb48cf7ade) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211556 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
e4efd9faf2 |
sw: compare positions, not pointers, when looking for a format redline
When a paragraph had a tracked format change before the selection,
formatting another author's tracked format change and rejecting the new
change kept the other author's formatting.
Since commit
|
||
|
Mike Kaganski
|
984efa1ff7 |
Simplify ChartExport::exportHatch
Make it fill the Hatch struct in all three cases, and only write it in one place, for consistency. Change-Id: Ibf4eb97e1c7046a5fe7397509fff9560140b5eef Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13683 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit f14a9a4d3d45b08da075aa5a34b8d7daa5cae6a4) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211597 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
67346d08e9 |
tools: keep the input deflate could not take in one pass
ZCodec::Compress read a fresh chunk over whatever deflate had left unconsumed. deflate stops as soon as its output buffer is full, and the write-back that frees that buffer only ran at the top of the next iteration, by which time the tail of the previous chunk was gone. Nothing said so. EndCompression answers with the number of bytes taken and no caller compared it, so the result was a shorter document than the one handed in, compressed perfectly well. Passing 100000 bytes through a 64 byte output buffer took 49152 of them. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I1044a46aa1d1f424e7d699a927e01a8427768919 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13208 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 23e2e9c2136d8b5c9fd6d9604bccd35e9b657571) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211555 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
a0c08a780e |
vcl: let PDFDocument list the objects an object stream holds
GetElements() hands out the tokens the file itself carries, so a caller hunting structure elements finds none of those a /Type/ObjStm holds - every one of them, in a file from a producer that compresses its objects. They are there already: Tokenize unpacks each object stream it passes and registers what it finds, so only an accessor was missing. The 64 places that scanned the token stream for an object ask GetObjects() instead. testTdf57423 and testStyleNamedStructureTypes compared the order as well. The map is keyed by object number, which for a structure tree is document order, where the token stream carried the emission order that writes a child before its parent - the type sequence reads "Document Table TR TD" now instead of ending "TR Table". Both hold the same 22 elements; testTdf57423 stopped indexing its figures and compares one string per kind. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I3e46e600e0aed3802f5ee2f8c6dc25a50002763d Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13207 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 50bf4f7acb344520ce2055bddfaf844c23239a8e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211450 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
0b3a0fae8c |
Related: tdf#166401 fix writing properties for empty placeholder
Regression after commit b41a780d139ec6ca5a5514db972ccf0e50058dfc (tdf#166401: support custom placeholder prompt text, 2026-08-17), where a placeholder that authored no prompt stopped writing text and got a bare <a:p/> instead. On a slide an empty presentation object reports no text through UNO anyway, so the paragraph the export wrote for such a placeholder never had a run. It carried the properties, and those are the formatting. In PowerPoint 2016 the original title showed white at 22pt, indented 56.5pt; the round-tripped one showed bold black at 17pt, indented 18pt. Write the paragraph the way a shape with no text already does: the paragraph properties, the run properties in endParaRPr, and no run. The prompt still stays out of the file. That branch of DrawingML::WriteText is now a method of its own, called from both places. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I034a4e14f388c90872069d32b303c6c143c1814c Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/10768 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit e2859f5b23f53281d80cbad4c56db1764aaf1406) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211538 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
8343c6e702 |
vcl: size an appended cross-reference stream's offsets by the file
WriteXRef wrote /W[1 3 1] and three offset bytes whatever the document. Signing a PDF larger than 16 MB recorded every offset in the appended update modulo 16777216, so the updated catalog and the signature itself pointed into the middle of the original body. The field is as wide as the largest offset needs now. /DecodeParms/Columns counts the bytes of a row apart from its predictor byte, so it follows the field rather than staying at 5, and each byte is shifted out of the offset rather than masked with an int too narrow to hold the mask. Reachable only from a file whose cross-reference is already a stream, which is why our own exports never met it. It has no test: the only path that appends a cross-reference section is signing, and that needs a certificate. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I08279fcb43d0c538f66dc029890a58a76f6dab7e Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13206 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 7e52e9eabbda9de67fedab5bf96a71902e53bc28) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211445 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
010611712f |
vcl: read a cross-reference stream that carries no predictor
PDFDocument sized every row of a /Type/XRef stream as one predictor byte plus the /W fields and read the fields from offset one, and the no-prediction arm of the filter never copied the row it had read. /Predictor defaults to 1, so a stream written without /DecodeParms decoded as nothing but free entries. It gave up entirely when /Filter was absent, which is as legal. An indirect /Length is resolved through those offsets. With none of them the tokenizer cannot skip a stream body, so it reads the compressed bytes as tokens and Read() fails on the file. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I4929e879e02ba91dce8280421f129f91dba9074d Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13205 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 3adab88af78016314f210f2be66668190b1f3493) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211444 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
59453af50d |
tdf#171555 sw: draw a non-breaking hyphen as a plain hyphen
Regression after commit 367ef7c1dca300b1966093de5bbfca8ea97d403a (tdf#167946: reimplement the fix for tdf#120677, 2026-02-11). A non-breaking hyphen was shown as an empty box on macOS, and on Windows in a font other than the rest of the text. Inside a field it wasn't painted at all, and the brackets around double-lined text came out as the text they enclose. SwTextFormatter::NewPortion builds the portion for U+2011 with a plain '-'. The SwTextSlot in SwExpandPortion carried that character to the font; dropping the base class sent U+2011 instead. Carlito, Caladea, Arial and Courier New have no glyph for it, nor for the U+2010 HarfBuzz falls back to. So declare the substitution in SwBlankPortion. A blank stands in for one character; it doesn't expand to computed text. Format, measure and paint all use that character where it differs from the model one; a no-break space is shown as itself, so tdf#120677 and tdf#167946 stay fixed. SwTextPortion::Paint splits in two for it, since the decorations read the index that the swap resets, and a blank with no length still has a width to paint. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I93af27f8647e7e04014734af7100372fa66018cb Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13203 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit e9ce5e4424deb3a28db630fdb540ccfac1f4c0a2) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211440 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
f0600a8d88 |
tdf#163272 sw: give a paragraph its own /Lang
Write a paragraph in a language other than the document's, then File - Export As - Export as PDF with Tagged PDF on. The paragraph's element said nothing about the language, and the Span holding its text said it instead: /StructElem /S/Standard /K[6 0 R] /StructElem /S/Span /Lang (en-GB) /K[1 2 3 4 5 6 7 8] A run of text was compared against the language of the document, so every run of a paragraph that declares one of its own differed from it and opened a Span. The comparison is against the paragraph now, and a paragraph unlike the document names its language itself. A run unlike its paragraph still opens a Span and names its own, which it did not where it was the document's language the paragraph had departed from. A Span that carries anything else - a strikeout, a script, a character style - still stands, without the language it was repeating, and so does a Link. A paragraph with something anchored in it says nothing, as before: what hangs there is tagged underneath it and would take the paragraph's language for its own. testSpans reads four list items, three of them en-GB in a de-DE document. It asserted the shape the old rule made - for one of them it looked for a single Span and took the language off that, and for the rest it counted the Spans that rule opened. It reads the language off the paragraph now, asserts that the item at the document's own language names none, and where an item holds a hyperlink and a footnote it finds the two Links alone. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I4082891666598793f59d1bf62c33e883ebdebb8f Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/13117 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 0d2d9e656d87647f904f1a463ed9b879c2bfb10c) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211375 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
5d6c85d78b |
vcl: a tagged PDF's lengths are in user space units
Indent a paragraph by half an inch, then File - Export As - Export as PDF with Universal Accessibility (PDF/UA) on. Its element read /A <</O/Layout/Placement/Block /StartIndent 0.72 where half an inch is 36 points. Every length was a fiftieth of what it names: a 64.8pt inline formula carried /Width 1.296 and a 481.9pt table /Width 9.638. Writer passes twips, and the PDF writer prints thousandths of a point. For an inline Formula the standard makes Width the sole source of information about the element's extent (ISO 32000-2 14.8.5.4.6), so the fiftieth is all a reader reflowing the file has. A bounding box was right all along, because SetStructureBoundingBox converts one out of the current map mode. The nine lengths this writer can set take the same conversion now. The counts and the identifiers pass through untouched - RowSpan, ColSpan, Language and the annotation ids. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I4da4d085de9da7e3773294756ff3205b9b1b38aa Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12997 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit ec0a83ea9cacd67b7842a4899e9996b2d0048ce0) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211350 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
f7bfdc2d03 |
tdf#164049 sw: a formula in a paragraph is not a block
Insert - OLE Object - Formula Object, write E=mc^2 inside a sentence, then File - Export As - Export as PDF with Universal Accessibility (PDF/UA) on. PAC 2026 puts 1 warning on the "Formula" structure elements row: Possibly inappropriate use of a "Formula" structure element. The element read /A<</O/Layout/Placement/Block/BBox[212.05 713.7 276.85 735.3]...>> Placement was Block for every frame but a table cell and a figure anchored in the body. A formula is tagged inside its anchor paragraph though - as character it sits in the text, and CheckReopenTag reopens the paragraph's tag for the other content anchorings - and ISO 32000-2 14.8.5.4.2, Table 378 reads "When applied to an ILSE, any value except Inline shall cause the element to be treated as a BLSE instead". Neither the description, the bounding box nor the size has a part in it, each ruled out one at a time in an exported file. The element reads /Placement/Inline now for a formula anchored as character, to the paragraph or to the character. One anchored to the page or to a frame is tagged under the document, where a block is what it is. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I57e5b52c5dd4e34a8588bc1091b13864e6d4497e Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12988 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit bae420ab335afaff185046486a3efbad59d6b150) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211349 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
c3bcf7d5f8 |
tdf#168720 sw: an undescribed formula's /Alt is its source
Insert - OLE Object - Formula Object, write E=mc^2, save and reopen, then export with File - Export As - Export as PDF and Universal Accessibility (PDF/UA) on. PAC 2026 fails it under Alternative Descriptions, the element saying nothing about what it holds: <</Type/StructElem/S/Formula /P 5 0 R /Pg 1 0 R ... /K[1 ]>> Writer takes a frame's Alt from the title and description a user typed; a formula has neither unless asked. A tagged export falls back to the formula's own source now, so the element says E=mc^2 and PAC passes that checkpoint: /Alt <FEFF0045003D006D0063005E0032> A title or description still wins. The report asks for an option; a formula with neither said nothing at all, so the source goes in with none to switch it off. Reading the source needs the object running, so the export runs each undescribed formula and puts it back. PAC's separate Structure tree warning still stands. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I09a68eff3d15981a7e112f79409b3448bceb4bc0 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12987 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 5b84c1873533bcdd98d917e8b6fb6d149a238047) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211348 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
d663a07822 |
sw,vcl: name the heading a TOC item lists
Insert - Table of Contents and Index, then export with File - Export as PDF, Universal Accessibility (PDF/UA) on and PDF version 2.0. The file fails ISO 14289-2 clause 8.2.5.8 once per entry, which veraPDF reports as: Each TOCI in the table of contents shall identify the target of the reference using the Ref entry, either directly on the TOCI structure element itself or on one of its child structure elements An item named no target at all: <</Type/StructElem/NS 2 0 R/S/TOCI /P 8 0 R /Pg 3 0 R /K[10 0 R ]>> It names the heading now, in the /Ref entry PDF 2.0 added: <</Type/StructElem/NS 2 0 R/S/TOCI /P 8 0 R /Pg 3 0 R /Ref[ 15 0 R ] /K[10 0 R ]>> Object 15 is the /S/H1 the entry links to, found from the mark its URL names. An item that links to a table, a frame or an object, or that has no link at all, still names nothing. The document of testTdf173162 linked its entries to bookmark alone now, as Writer writes them, and the document passes the profile, so the test asks for PDF/UA instead of plain tagging and the harness validates it. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I76e73684d2ab09f187f56cab093d13980150d737 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12750 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit ee49bf58e7d3eb3ff94677e9a51c73b76eb51a99) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211338 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
07ccc05cf7 |
sw,vcl: name a footnote FENote at PDF 2.0
Put a footnote in a document and export with File - Export as PDF, Tagged PDF on and PDF version 2.0. The file fails ISO 14289-2 clause 8.2.5.14 once per footnote. veraPDF, on the PDF/UA-2 profile, reports: The Note standard structure type shall not be present in conforming documents unless role mapped to a structure element in the PDF 2.0 namespace The element carried the type PDF 2.0 replaced, in the 1.7 namespace ISO 14289-2 8.2.4 accepts alongside the 2.0 one: <</Type/StructElem/NS 2 0 R/S/Note /ID (id29) /P 28 0 R /Pg 3 0 R ...>> Writer asks for FENote now, so the element lands in the 2.0 namespace and the check stops failing: <</Type/StructElem/NS 1 0 R/S/FENote /ID (id29) /P 28 0 R /Pg 3 0 R ...>> Below PDF 2.0 the file still says Note, the only footnote type ISO 32000-1 has. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I4674f4f902621ee8765121782799f42cc59c5e6e Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12749 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 42433cf89b23328991e7a64daea4f771b60f728a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211337 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
3e517fbfe9 |
vcl: give each style a structure name of its own
Write a paragraph in Writer's built-in Index paragraph style and export with Tagged PDF switched on. The element comes out as /S/Index, over a role map that reads /Index/P so the standard Index type means a paragraph, document-wide. veraPDF, on the PDF/UA-1 profile: Standard tags defined in ISO 32000-1:2008, 14.8.4, shall not be remapped A style called Artifact went further. Its paragraph reached the content stream beside the marker for content that is not part of the structure at all, the two being one name: /Artifact BMC /Artifact<</MCID 10>>BDC Writer names a structure element after the style that produced it, and the role map records what that name means, one entry per name. Styles collide there with the standard types, with each other, being separate families for paragraphs and characters, and with the names Writer asks for itself, Source Text writing Code literally. A name a standard type has, or another alias asked for, is now passed over for the first indexed name beside it that is free, so the Index style writes /S/Index-1 over /Index-1/P. PDF 2.0 has standard names 1.7 lacks: a style called H7 gives way there, while a heading at outline level 7, which Writer asks for by that same name, keeps it and needs no entry at all. The synthetic Div containers ask for no name now, and carry the Div tag as before. The test document remapped four standard tags at PDF 1.7 and six at 2.0; now none at either. Found while reproducing tdf#165454. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I63e9116e5db17bdb554b508a864e313a77f40bc2 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12361 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 54e0207ff6bcf937187a395912f1ed57b60d90e3) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211314 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
cfbbd638ab |
tdf#168638 sc: tag a database range's labels as TH
Mark a table with Data - Define Range and the option "Contains column labels", and export with File - Export as PDF and Tagged PDF on. Every cell of the range comes out the same: <</Type/StructElem/S/TD /P 7 0 R /Pg 1 0 R /K[9 0 R ]>> so nothing in the file marks the row that names the columns. The label cells are TH now: <</Type/StructElem/S/TH /P 7 0 R /Pg 1 0 R /A <</O/Table/Scope/Column>> /K[9 0 R ]>> A range sorted left to right holds its labels in the first column and gets /Scope/Row there. A range Calc defines itself for a sort or an AutoFilter counts the same, on the header row it guessed. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I5f4d3ec709f330604d518e1bf0c89c1655b18816 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12657 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 8a524317421379ba257fcfba04f63f06da7cdcc1) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211174 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
fe55fbaa27 |
sc: tag a sheet the page range starts inside
Export a two-sheet spreadsheet with File - Export as PDF and a page range that leaves out the second sheet's first page - "1;3" where that sheet spans pages 2 and 3. The file gets one bookmark where it should get two, the second sheet having none. Switch Tagged PDF on as well and it also carries one Worksheet element, holding both sheets: 5 0 obj<</Type/StructElem/S/Worksheet/P 4 0 R/Pg 1 0 R/K[6 0 R 22 0 R]>> where 6 is the first sheet's table and 22 the second sheet's. The per-sheet work ran on nRenderer == nTabStart || bIsFirstPage but the renderer index is the page's own index in the document and a page outside the range is never rendered at all, so a sheet whose first page is left out never reaches it. It asks whether the sheet changed since the last rendered page now. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I158feb5ca4a6f4501e6dfdf5234c8fdf75b8ee40 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12571 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 49e441a059f98d703ce2bddd5b0eabf80b7144aa) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211173 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
5c7062bde2 |
tdf#171022 sc: name the element a cell's hyperlink points at
Give a cell a link to another sheet with Insert - Hyperlink. Export
it with File - Export as PDF, Tagged PDF on. veraPDF, on the PDF/UA-2
profile, fails clause 8.8 once per such link:
Destination in Outline item, OpenAction or Link annotation is not a
structure destination
The link annotation named a page and an area on it, and that is all:
/Dest[28 0 R/XYZ 56.693 445.607 0]
I56c2f517977d4b876e463f4bcadc53f8ae444b08 ("tdf#171022 sc: name the
element an outline item points at") gave the export a list of
destinations to name an element for, resolved once the last page is
drawn. A hyperlink goes in that list too. The export now keeps each
cell's element as the cell is drawn, so a hyperlink onto a cell names
that:
/A<</Type/Action/S/GoTo/D[28 0 R/XYZ 56.693 445.607 0]/SD[33 0 R/XYZ 56.693 445.607 0]>>
/SD names an element where /D names a page. Object 33 is that cell's
/S/TD, and its /Pg is that page. A sheet's /S/Worksheet has the
sheet's first page for its /Pg, so it can stand in only for a
destination there. A link onto a cell nothing was drawn for names the
sheet then, and keeps the plain destination on a later page. So does
a link to a sheet that prints nothing. The test document goes from
four failures of the clause to none.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: I3b1dc51e9962ac19061934d447da9542ee56b0fd
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12393
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit 51a84acd2fccf7120959ff7ff16a0549ddb34f58)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211172
Tested-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
9ec63cdce0 |
tdf#171022 sc: name the element an outline item points at
With Tagged PDF and Export outlines both on, export a two-sheet
spreadsheet with File - Export as PDF. veraPDF, on the PDF/UA-2
profile, gives two failures of clause 8.8, one per sheet:
Destination in Outline item, OpenAction or Link annotation is not a
structure destination
Each sheet's outline item named a page and an area on it, and that is
all:
/Dest[1 0 R/XYZ 0 792 0]
Calc creates the outline item before drawing the sheet, and opens the
Worksheet element only while drawing. So the destination can't name
the element where it is made. The export state now outlives one
sheet: it keeps each sheet's element, and each destination with the
sheet it points at. The last page joins the two:
/A<</Type/Action/S/GoTo/D[1 0 R/XYZ 0 792 0]/SD[5 0 R/XYZ 0 792 0]>>
Object 5 is the first sheet's /S/Worksheet, and the clause passes.
The Writer half is Ic76d80d46f130d549ebaacf175393a3d6ec77998
("tdf#171022 sw,vcl: name the element a link points at"), which added
the vcl call used here. A hyperlink to a cell still gets a plain page
destination, and needs its own change.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: I56c2f517977d4b876e463f4bcadc53f8ae444b08
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12366
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit 3571fc6be8f8e33ce25a1e83ac4d5dcd8da6494a)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211171
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
c3cae7d482 |
drawinglayer,svx: ask for an artifact instead of faking a tag
I9431c460cade3d432cb7abf76930e23f8106a7f1 ("tdf#168462 svx: mark the
slide background as an artifact") left page decoration reaching its
/Artifact by a lie: VclMetafileProcessor2D emitted one only for a
background image, and dropped a tag naming NonStructElement, so the
tag claimed to be a Division and an image. Decoration now says so
itself: the artifact decision reads the decorative flag, and the
caller passes NonStructElement and no image and means them. A
background image keeps its own path to an artifact.
isTaggedSdrObject() becomes opensStructureElement(); its comment
still claimed every one of these primitives comes from an SdrObject.
The exported file is unchanged. The one call that goes away with the
old predicate, alternative text on a tag that opens an artifact,
could not reach the file either: setAlternateText writes nothing
without m_bEmitStructure, and checkEmitStructure() is false inside a
NonStructElement. operator== now compares the element and the
decorative flag, which decide what comes out and were not compared.
Letting a master slide's own shapes out as artifacts does not follow
from treating every background as decoration. A background object can
hold tagged children, and a Paragraph begun inside an artifact opens
no marked content at all; tried on testTdf135192, three pieces of its
content came out in no marked-content section. Something narrower has
to tell the two apart.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: Id0fc0d9bee36e8182702d77e9ac9647709a63f68
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12321
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit 9353a1cac3e8056d175acc29dbeecc79ff1cdcd4)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211170
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
f1e4a0f0ab |
tdf#168462 svx: mark the slide background as an artifact
A slide with a background, its own or its master's, exported to
tagged PDF with that background as untagged content. ISO 14289-1 7.1
leaves no third state, and veraPDF says so once per background:
Content is neither marked as Artifact nor tagged as real content
Page decoration is wrapped as an artifact in
ViewObjectContact::getPrimitive2DSequence, which the background never
reached: the master page descriptor builds its own hierarchy and
emitted the fill straight into the visitor. It now takes the same
wrapper as the page decoration beside it - the machinery of commit
|
||
|
Mike Kaganski
|
32dec34df6 |
tdf#171022 sw,vcl: name the element a link points at
Exporting a document to PDF/UA-2 produced a file that fails ISO 14289-2 clause 8.8, once per outline item and internal link. veraPDF reports Destination in Outline item, OpenAction or Link annotation is not a structure destination The destination named the page and an area on it, and that is all: what was linked to could only be worked out by looking at what the layout had put at that position. A destination for a link or an outline item is now written as a GoTo action carrying a structure destination beside the page one (ISO 32000-2 12.3.2.3): /A<</Type/Action/S/GoTo/D[<page> 0 R/XYZ ...]/SD[<elem> 0 R/XYZ ...]>> sw creates a destination before its target is tagged, so the node is remembered while the cursor is on it, and matched against the tags in the pass that already runs after the last page. Creating one and remembering its node is a single call now, which every destination a link or an outline item points at goes through: bibliography entries and links in the text of a shape are named as well. The cursor is not always on the target: a jump that selects a frame or a drawing object, and one that finds nothing, move it nowhere, and where it was left standing on the link, that link's own element must not become the target. The pass over a shape's links can instead start with the cursor on the target already, where a jump that did reach it shows no move either, so there the mark names its own node. Those keep writing /Dest, as do a target that was never tagged and one whose element was never emitted. The element carries its role and its content, not a name of its own: a heading becomes /S/H1 with no /Alt or /T. veraPDF does not check what a structure destination points at, so the test does. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ic76d80d46f130d549ebaacf175393a3d6ec77998 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12300 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit a6502c6e96d13425bcb4a3bf5731a998982d5957) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211150 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
6e785d7b0b |
tdf#163240 sw: subdivide a Link by Spans
The commit "tdf#163240 vcl,sw: keep one Link tag for a wrapped link" made language, baseline shift and decoration come from the link's first portion for the whole link, and noted subdividing by Spans as the fix: those three are attributes of the Link tag, and there is now one tag per hyperlink, so a change inside it had nowhere to go. Open a Span in the Link instead. A span waits inside its link, so two tags can be open at once, and an annotation goes to the link's own tag - vcl takes it on a Link only, and a styled run crossing a line break would otherwise leave that line's annotation without a StructParent. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I302d32f66fcb4e3b1eaf63b136bdcf96aaea17b7 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/11543 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 8f52d63f897aaf710dfe4cd9771f39689471abfd) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211101 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
6e70cb2e00 |
tdf#163240 sw: track which tags to close, not how many
The commit "tdf#163240 vcl,sw: keep one Link tag for a wrapped link" kept the tag across a line break. A ruby still split it, and a character style on a ruby base pulled the following text into the Ruby element. Tags were closed by a count, which callers decremented to leave one open for the next portion; a count does not record which tag. Keep two stacks instead: the tags this helper opened, and the tags left open for a following portion. Close a deferred tag only while it is the current element, so neither a ruby nor a fly closes a tag of another. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Icdf793f149076db2b5dfac2b7668d0c414467427 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/11542 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 1936ac2268b9e603c64a03eff6eb8b7ecb8b56de) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211090 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
79b06ca748 |
tdf#163240 vcl,sw: keep one Link tag for a wrapped link
Regression after commit
|
||
|
Mike Kaganski
|
71a7e22201 |
tdf#173162 vcl: put dropped types in the PDF 1.7 namespace
Exporting a document with a table of contents to PDF/UA-2 produced a
file that fails ISO 14289-2 clause 8.2.4, once for the TOC element
and once for each entry in it. veraPDF reports
Non-standard structure type http://iso.org/pdf2/ssn:/TOC is not
mapped to a standard type
Non-standard structure type http://iso.org/pdf2/ssn:/TOCI is not
mapped to a standard type
Commit
|
||
|
Mike Kaganski
|
fd583f101a |
tdf#166963 vcl: reparent the children whose kids a split moves
A document with a table of more than 8191 rows, exported to tagged PDF, failed PAC with 1393 of Structure element refers to a different parent than the actual parent The kid list is split into generated Div containers. But m_aChildren and m_aKids aren't parallel: a non-structure element gets a child entry and no kid, and marked content adds a kid with no child. Advancing both by the same count moved the wrong children. Cut the child list where the moved kids end instead. addInternalStructureContainer also held a reference into m_aStructure across an emplace_back on it, so a split that reallocated left it dangling. It takes an index now. The test needs a document past that threshold. It takes about two minutes locally, most of it in the PDF parser it reads the result back with. Disable the test if that costs CI too much. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I37cb4f3e187b8a836b26e7c3da802bd12147904e Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12187 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 901e5afec874f0b16959038425365198dfbac7e7) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211066 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
dfdee8ff8a |
tdf#173194 sw: export a derived Table Heading as TH
A table cell whose first paragraph uses a style inherited from Table Heading was tagged TD, so a screen reader announces it as data rather than as a header; only the built-in style itself became TH. lcl_IsHeadlineCell compared the paragraph's own style name and stopped there. Walk the DerivedFrom() chain, the way the HTML table writer already does. Scope stays Column on every TH, so a row header still claims its column; that is a defect of its own. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I9ccbd1b3837d3edc5087162a143560e8412a846d Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12168 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 270d087cad2c8327775ce204371830c34eecc0d7) (cherry picked from commit d86b837d439355e62afcdd6b771603c260bc5949) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211049 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
7f398acee4 |
tdf#173217 editeng,xmloff,oox: keep a hyperlink's name
A hyperlink in Impress or Draw can carry a name, the Hyperlink
dialog's Name field: the description a screen reader announces, and
what PDF export writes as the link annotation's Contents. Saving the
document and loading it again lost it, and a tooltip carried by a
PPTX never arrived at all, so the link fell back to describing itself
by its own visible text.
Commit
|
||
|
Mike Kaganski
|
ab5bdc9cdd |
Use IsTextNode instead of comparison of GetNodeType result
Looks more consistent to me. Change-Id: Id61cb07ece99766b31713006ad3cc4d46c362791 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12304 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 1c7834da4d12de8458d57eee25e2fe97b5c6ace5) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211047 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
1eca0c353a |
Simplify SwFillCursorPos ctor using default initializers
Simplifies code, no functional changes. Change-Id: I433a3aa697d8bf3056d51894137f783821915970 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12051 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit ce897ee2f106bf890b8e3d431e0e34f48c9d9c65) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211046 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
121c5f332f |
Drop SwSpecialPos ctor, and use default initializers instead
Simplifies the code a bit, no functional changes. Change-Id: I72f71db5e12945d904b8922987bf7d5a4d231481 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12050 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit 72bfe9ed3e36faf927a31e235fef3fb91ae52851) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211045 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
7c2c72d189 |
tdf#167121 sw: tag the link an index mark covers
A hyperlink whose text an alphabetical index entry covers exported a link annotation with no tag around it, so PDF/UA checkers reject the file: Link annotation is not nested inside a Link structure element that is ISO 14289-1 7.18.5. The same happens under a reference mark or an RDF metadata field. Each of the three lays its run out as its own portion type, a SwTextPortion whose only addition is a shading that never reaches an export. BeginInlineStructureElements listed Text, Lay and Para, so the run matched no case and got no element at all - not even the Span carrying its language. Handle the three with the other text portions. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ie38193f64923cd6a0a3afa509691fb29a007f272 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/12104 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 409e0cc615e191a0392ee080717bb84eb61e4724) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/211027 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
32ab4a61cc |
Simplify SwPosition comparison operators
Drop inequality, that is automatically deduced; and use less-than to define others. Change-Id: Ib979cbe74ec28c1fde0337276d4f602a1fe907b4 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/10674 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit 9f8156beb5eb02ef7d7b00a3effbd6631fc58dfb) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210507 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
2141c9e689 |
Simplify some SwNodeIndex methods
Delegate where possible, and drop operator !=. Change-Id: I64efc13539dc78c4858775c10d3287dc416467fe Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/10733 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit c4a06b011473941dfa0b258f3ba655e7f1a9cc3e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210489 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
f56b27b9a0 |
cool#16080: give the PPTX slide master the shapes its layouts share
If6c23baae18f1d4bd18421a9e48da74956e4a47a ("cool#16080: keep a layout's shapes
off the slide master") made the master part write only placeholders, so a
master that drew anything of its own lost it and every layout kept a copy.
Each page of a group draws the master's shapes first and its own after, so the
leading run they all have alike is the master's: write those on the master and
leave them off the layouts. A shape only one layout draws is not in that run,
so it stays where that commit put it.
Comparing what the pages have in common, rather than where they came from,
works for a deck authored in ODF too. Two shapes are the same when what the
export writes for them is the same; what a page recalculates - a frame's
laid-out rectangle, a cached ordinal number - is left out of it.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: If0660def98c52bbbade6b747c926806d7c4064b8
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/10138
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit d27410878eba5dc16d601e97a2e0d5e2bed1399c)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210365
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
b51d976d0a |
cool#16080: keep a PPTX layout's shapes off the slide master
An imported master and layout collapse onto one Impress master page, so the page standing for a group of layouts holds that layout's own artwork; writing its whole shape tree into the master part painted a full-slide rectangle from the title layout under every slide. Write only the placeholders there - a placeholder reaches a slide only through a layout that references it, a drawn shape paints unconditionally - and let each layout write its page's shapes. A master of its own keeps writing them, since its layouts are written with no content. A master carrying artwork of its own now loses it from the master part while every layout keeps it; telling those apart needs the shapes common to the whole group. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: If6c23baae18f1d4bd18421a9e48da74956e4a47a Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/10086 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 9928c7a8badb4b20fded7a18ba4a59fc0aba44f7) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210364 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
cd1536dfd8 |
Related: tdf#167713 sw: keep a replayed shape out of the row buffer
Converting a document whose header holds a table with a shape in a cell
aborted the RTF import in a Debug CODA build:
C:\PROGRA~1\MICROS~1\18\COMMUN~1\VC\Tools\MSVC\1451~1.362\Include\deque(1128) : Assertion failed: back() called on empty deque
reached, as crashtesting found it, with
COConvert.exe rtf fdo35334-3.odt
COConvert.exe pdf fdo35334-3.rtf
on https://bugs.documentfoundation.org/attachment.cgi?id=57398.
The row buffer is replayed while the states still point at it, so the shape
insertion was buffered again, behind the cell being replayed, and dropped
with the buffer; the EndShape next to it then closed a shape that was never
opened, taking the table manager that the next substream needed.
Regression from 6c7c171d1468765f7e2a1872d7243f6bf809e37e (tdf#167713 sw:
keep an RTF pict shape in its cell, 2026-08-06).
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: Ib4fe247910cf67a8b60d448da75510bcf2251885
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/10084
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
(cherry picked from commit b2a7dd57076573ddac7141b99f9c33a68fbe799d)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210363
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
26d693d85e |
sd: extend Undo to cover the placeholder mechanics
Deleting all the text of a placeholder was not undoable: sd puts the prompt back before svx captures the text change, so the action it creates sees the prompt as both the old and the new text and drops itself. The text is now recorded before the prompt goes in, in one group with the change of what represents a picture placeholder, so one step takes the whole deletion back. Two things that only surface through that step: a kind with no style sheet of its own kept none and painted itself in the pool default fill colour, and the placeholder built to stand in for the text recorded an insert that redo performed a second time, on top of the swap. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I7bd581216b9111197144d4691c2a12c996cf0e3f Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9442 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 33b95acb2e26a22d3242605f7d609108a678620e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210330 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
4c9116ee9d |
tdf#166401: keep text in a picture placeholder
A slide's picture placeholder can hold text the author typed instead of an image, which PowerPoint displays. Impress dropped that text on import and wrote nothing at all for such a shape on save. To reproduce, type into a picture placeholder in PowerPoint, save, open the file and save it again: the text is gone from the shape, and the shape from the file. What a placeholder is now lives in the page's presentation object list, and what represents it follows what it holds: a graphic object for a picture, an SdrObjKind::OutlineText object while it holds text, the placeholder itself while it holds nothing. It changes over when a file states such text, when that text is deleted again and the prompt comes back, and when an image is put in. Text arriving another way, through the API, is caught by the new SdrPage::onEmptyPresObjFilled and switched at the next idle, so that route needs a spinning scheduler; what a save writes is right either way. An object in text edit is left to that edit, which is where it changes over: the idle pass walks every page, so it asks each object rather than trusting the one that asked for the pass - in a shared document another view holds it. That switch is the model's own and covered by no undo step - it answers for a document the API changed, not for anything the user did. What the file says about the shape travels with it both ways, the graphic items included, so an outline a picture placeholder is clipped to is still there when a picture arrives. It is not written while the placeholder holds text: the file says text box, and a text box states no geometry to clip a picture to. PPTX writes it as the placeholder it is, with the text it holds. ODF states the class of what represents it, so the placeholder's own class travels beside it as coext:placeholder-class, and a reader that ignores the extension sees an outline placeholder with the text. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I0f3353115e56d63a4158ebaeb5689689499894bb Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9441 Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 7921401ae48188568e1016db00bd90256d666a6a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210329 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
3d01685a4f |
sd: fix crash setting IsEmptyPresentationObject twice
Making a presentation object empty took the style for its prompt from the first
paragraph of the text it holds, without checking that it holds any: a graphic or
an OLE one never does, and making an object non-empty removes what text it had.
There is no message to quote - SdXShape::SetEmptyPresObj dereferences a null
OutlinerParaObject and the process goes down where it stands. To reach it in a
build without this fix, open a document whose slide holds an empty picture
placeholder, such as sd/qa/unit/data/pptx/tdfpictureplaceholder.pptx, and run in
Basic:
oShape = ThisComponent.DrawPages(0).getByIndex(0)
oShape.IsEmptyPresentationObject = False
oShape.IsEmptyPresentationObject = True
Which way the prompt runs was taken from the text as well, so an object without
any got a horizontal one; it answers for that itself now.
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: Ie5ba7ee52e1e9a2131dbd09e1df4c773ea52431c
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9304
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
(cherry picked from commit a8df64888cad38c406929cd5a63960cdb8fdc545)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210325
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
||
|
Mike Kaganski
|
84c8ca6fd5 |
Address review comments
See https://gerrit.collaboraoffice.com/c/online/+/9927 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I6139405c1285914323a4c2e0b3ef6ace0034195b Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9953 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 609c2a17a654452409ec090c47351b49c4990349) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210335 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
ecf0efa073 |
tdf#165609: save the form control label font and alignment
The XLSX export wrote the label as a bare <font> element, so a checkbox came back with the control default font at 10pt and adjacent labels overlapped. The label's font name, size and colour, its bold, italic, underline and strikeout, and its horizontal and vertical alignment are now written out. The underline stopped reaching the export altogether when the control property list started asking for "FontLineStyle", which no control model has: commit |
||
|
Mike Kaganski
|
64fe896867 |
tdf#166401: keep a placeholder's prompt in ODF
Commit
|
||
|
Mike Kaganski
|
72b866f367 |
cool#16083: an empty placeholder stays empty
Formatting an empty presentation object wrote the outliner's emptiness back over the prompt the object holds, leaving a blank frame with nothing to insert into. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I7c168196dfccfb5d0df9b01cc9d4c52c62db5555 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9287 Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit d4aa2b9e28adbe7bb241bdee33c5af91e85258b4) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210236 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
5f26da385c |
cool#16083: do not take a prompt as slide content
A slide placeholder took its layout placeholder's paragraphs as its own text, so it arrived filled and stopped offering to insert a table, a chart or a picture. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I879dc64e0129101623fda42361afb13b6be3e4e3 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9286 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit e00c44aa81d7d695a34eddf7e672f926aa7fc955) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210246 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
95cbf113ed |
cool#16082: keep the subtitle a slide layout holds
A subtitle placeholder was written only for a page that is not a master, since commit |
||
|
Mike Kaganski
|
9b3b1373b2 |
cool#16080: write the body the slide master holds
The master part was written from the page standing for the group of equivalent masters, so it carried that layout's placeholders and nothing else - for a deck whose first layout is a title slide, no body at all, and PowerPoint offered no Master text styles to edit. Take a type a master may hold that this page has none of from the first page of the group that has it. A layout placeholder was imported with what it inherits from the master, so the first layout with a body carries the master's own geometry, which is where PowerPoint puts the box. A deck whose first body-bearing layout states a geometry of its own hands that over instead. testTdf106867 moved its expected spid from 65 to 66: the master part writes one shape more, and shape ids are document-wide. It still names the p:pic that holds the video. Underneath is a document model difference: PowerPoint has masters that carry layouts, Impress has masters alone, so the import maps each layout to a master of its own and the export reconstructs both levels from them. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I1e6d311e5f20cd3377ce4c0c3f2dfa59ff4dfefb Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9196 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 73dfd24f01a6fdf5fbd7c0d48f17c027442fb870) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210234 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
605ce703de |
cool#16080: import the layouts in the master's order
The layouts came from the slide master's relations, which carry no order: a relationship id is opaque, and a producer may assign any id to any part. The map is keyed by that id as a string, which is what made it visible - a master with ten or more layouts handed over rId1, rId10, rId11, rId2 ... and the master pages kept that order. Read the master's <p:sldLayoutIdLst> instead, which is the order it lists them in. Beside the layout order itself, this decides which master page stands for the group of equivalent masters: before the fix, a second save picked the page of a layout with no title, and wrote a slide master without one. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I6ff62fd190742246a8317f9efae907eeb4a9d8a3 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9194 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 5652d7d68358aae8819ade918c5f5c3cf35f5456) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210143 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
2391b3dec0 |
cool#16080: keep the slide master's own shapes
A master standing for a group of equivalent masters wrote a shape tree with nothing in it and left the content to its layouts, so PowerPoint showed an empty slide master: nothing for a layout to inherit, nowhere to make an edit meant for all of them, and master level header and footer state that PowerPoint refuses to set while the placeholder is missing. Write its shapes into the master part as well, the way PowerPoint authors one. A placeholder index now belongs to the shape rather than to whichever part writes it first, because the master part and the layout write the same shapes. Without that a slide named the master's indexes while its layout carried fresh ones, and a slide pairs with its layout by type and index - PowerPoint's own files pair those two and number the master separately. testTdf106867 had to move its expected spid from 61 to 65: the master part writes four more shapes and shape ids are document-wide, so the animation's target moved with them. It still names the <p:pic> that holds the video. Underneath is a document model difference: PowerPoint has masters that carry layouts, Impress has masters alone, so the import maps each layout to a master of its own and the export reconstructs both levels from them. Two limits stay, both from an imported master and layout collapsing onto one Impress master page: the master part is written from the page that represents the group, so it carries that layout's title geometry rather than the master's own, and there is no body placeholder left to write. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I4c1465718918886ecfda16baafcd803e4cb095fb Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9193 Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 270d83101f69bdfe8980a14414d08f88b64232ca) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210142 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
4bfb5aaa80 |
cool#16079: keep the date field in a placeholder
A date/time placeholder saved to PPTX held our own display string "<date/time>" as literal text with no <a:fld> at all, so PowerPoint showed that string where the original showed a date. DrawingML::GetFieldValue has no case for the field kind a header/footer date carries, "DateTime", and with no type name nothing writes the field. Name the type that leaves the format to the reading application, datetime: the field carries no format of its own, the page holds it. PowerPoint writes datetimeFigureOut in such a placeholder, an alias of datetime1 - MS-OI29500 part 1, 21.1.2.2.4 (fld) - which would pin a formatless field to MM/DD/YYYY; that alias is read on import as the datetime1 it stands for, so a slide's own date field keeps its format across a round-trip. Every other datetime* type was understood; the one PowerPoint itself writes was being dropped. The page's own DateTimeFormat is out of reach where the type is chosen, so a footer set to show a time is written as a date. PowerPoint has no per-page format to fall back on either way, and a date beats the literal string this replaces. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Icbd9ac57ed7b2311d5f6570fffad1cd78e4d6bf7 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9192 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 145a37d84dd3b22023c46b6be8577008f36a48b9) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210141 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
1dc3b23162 |
cool#16078: OOXML round-trip of the clip polygon
A placeholder clipped to an outline carries it as custom geometry, which an image shape has nowhere to keep, so the outline was dropped on load and the save wrote a plain rectangle over it. Read it into GraphicClipPolyPolygon, and write it back from there. The two directions are not equally capable. Import reads paths of literal coordinates: one driven by guides or formulas, and an arc, both need the custom-shape machinery. A quadratic curve is read, converting exactly to the cubic the model keeps. A path that states no w and h is not read either - such coordinates are absolute EMU rather than a space to scale from, and no file authored that way turned up to check the reading against. Export writes whatever polygon the model holds, so that direction loses nothing - a geometry the import gave up on was never stored. A curve that closes a contour states no end point of its own, and the point the contour started from answers for it. A path that states fill="none" states that it paints nothing, and PowerPoint leaves it out of the clip: a picture clipped by two paths shows through the plain one and not through the nofill one, while the same file with that one attribute removed - or with stroke="0" in its place - shows through both. So such a path is read past, while the shading variants and stroke="0" are not: those say how an area is painted, not whether there is one. Several paths become one polygon by area, not by concatenation: the contours of one path are filled by the even-odd rule, so two of them overlapping leave a hole, while paths are filled one after another and the shape covers what they cover together. Concatenating them instead cut a hole wherever two paths overlap, which the file does not describe and PowerPoint does not draw. A shape's own custom geometry now replaces the one it inherits, the way a preset geometry already did. Saving states the outline on the slide as well as on its layout, so reading that back gave the slide placeholder its own geometry beside the inherited one, and two geometries at once are not one this reads - the placeholder came back unclipped while PowerPoint showed the same file correctly. The outline reaches the drawing the commit before this one added, so a placeholder read from a file shows what sits behind the parts it does not cover. It survives a round-trip through ODF as well, where the property already travels as coext:graphic-clip-poly, with nothing to carry until now. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I87bd21e68bcea648c8c9151b6af5439578f1c2cf Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9303 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 04bbb74a24bad4ae54f257b6b028f3fd48955272) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210123 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
be205bb214 |
cool#16078: ODF round-trip of the clip polygon
Carry the GraphicClipPolyPolygon UNO property on graphic-object shapes through ODF save/reload as a new <coext:graphic-clip-poly> child of <draw:frame> with the polygon serialised as svg:d (the same SVG d-path encoding draw:path uses). A new XML_GRAPHIC_CLIP_POLY token is added in lockstep across xmltoken.hxx, xmltoken.cxx, and tokens.txt so the XMLTokenEnum and FastParser tables stay aligned. The element is taken before the frame decides what its shape child is, and applied when the frame ends: an empty presentation object has no draw:image for the polygon to follow, so the polygon comes first and the shape it belongs on does not exist until then. A placeholder saved to ODF came back unclipped otherwise, which is the shape of file our own save writes. It goes on the image the frame keeps, which is not always the last one it saw: a frame with a replacement image holds several, and the ones that lose are taken off the page and disposed. Export goes through a new ImpExportGraphicClipPoly helper called from ImpExportGraphicObjectShape alongside ImpExportSignatureLine and ImpExportQRCode (under ODFSVER_EXTENDED). Import is handled in SdXMLFrameShapeContext::createFastChildContext, which is where <draw:frame> children land - the GraphicObjectShape sub-context is nested inside that and never sees this sibling directly. The element is added to the extended ODF schema next to loext-signatureline and loext-qrcode, so a document carrying it validates, and is listed in the Collabora ODF extensions notes. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ib9ae65c1b6f2d8fdb28ae086fd7e2be2f8a8ee24 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9191 Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit b7a56041afda6b658b6b16bb51a5939295c40d75) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210122 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
e26a668059 |
Avoid useless call to CreateColorBitmap
It is potentially expensive, and isn't needed: the ctor of Bitmap taking a bitmap and an apha mask only uses colors from source bitmap anyway, not its aplha. Change-Id: I086fec94659495020108dc88a75b5479aff53b66 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9533 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit a16c260b2729bfc64a8d2cc9ae74c99f978bb506) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210124 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
dec2f2b9ef |
cool#16078: draw a graphic object's clip polygon
The clip a graphic object carries was kept in the model and carried through save and load, but nothing ever drew it, so a shape a file describes as an outline still covered its whole frame and hid what sat behind the parts the outline leaves out. Wrap the shape's content in a mask. The polygon states the shape's own coordinates, so it is normalised and then put through the same object matrix as the content, and follows the shape's rotation and shear along with it. A flip is expressed on the graphic itself rather than in that matrix, so the clip is mirrored here to follow it - the picture would otherwise turn round inside an outline that stayed where it was. The mask covers the fill as well as the graphic, which is what uncovers the background. The invisible outline stays outside it, leaving bounds and hit-testing as they were. A mask clips with an even-odd fill, which is what makes a contour overlapping another one a hole. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I69c8c6c31f43e87399fb582665904dc81221b32b Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9302 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 120b9ef0b7c50a0e025e7be87e005aa39ab7db8a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210037 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
9d36b250d9 |
cool#16078: add SdrGrafClipPolyPolygonItem
Introduce an SfxPoolItem in the SDR graphic attribute range that carries a basegfx::B2DPolyPolygon clip outline on SdrGrafObj. Allocate SDRATTR_GRAFCLIPPOLYPOLYGON between SDRATTR_GRAFCROP and SDRATTR_GRAF_LAST (which shifts SDRATTR_3D_FIRST and the downstream ids by one), register the pool default, and surface the value as the "GraphicClipPolyPolygon" UNO property on the GraphicObjectShape property map (covering both drawing.GraphicObjectShape and presentation.GraphicObjectShape via SVXMAP_GRAPHICOBJECT) using the css::drawing::PolyPolygonBezierCoords type. QueryValue/PutValue do the round-trip through the existing basegfx helpers. The new which-id is 1175. The trailing comments in svddef.hxx have run 2 high since two XATTR items were dropped upstream, so the ones added here say 1177 to stay consistent with the file. No reader, writer or render path consumes the property yet; this is infrastructure for upcoming custGeom round-trip work on empty picture placeholders. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I17d98cabeb230c25b3de0da4eadd8338ed672ce1 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9190 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit be7f21fc7187bb87129eed4e8ed2f794faf6e757) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210036 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
9fac5940ce |
cool#16078: write a schema-valid empty draw:image
A graphic-object shape with no embedded graphic data exported as <draw:image><text:p/></draw:image>. Per ODF's draw-image content model draw:image has a choice between common-draw-data-attlist (the xlink:href and friends attributes) and office-binary-data, followed by draw-text (text:p etc.). Without either alternative of the choice the element is incomplete, and the validator rejects it - the diagnostic mentions text:p because the parser stops at the missing choice content before reaching the draw-text part where text:p would be valid. Emit an empty <office:binary-data/> placeholder when there is no xlink:href and no graphic data, satisfying the choice. Groundwork for the clip polygon that follows, which is written on exactly such a shape: a picture placeholder that never got an image. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I5bf792c60affd35b33779c91be0a381eb074f00f Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9189 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 6d1cbfc9fb8e9a705b05f6cdddd4a87a630e973d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210035 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
c4256a2a76 |
cool#15754: round-trip empty media placeholders
Commit 73f1e76b1760d4e936aa22627623c922dcda964b (cool#15754: Fix PPTX round-trip of empty picture placeholders, 2026-05-21) covered <p:ph type="pic"> only. A layout's media placeholder still imported as an outliner shape, and an empty slide-side one exported as a <p:pic> whose <a:videoFile> named an empty external relationship - a package the Open XML SDK validator refuses to open at all. A media placeholder needs no copy of its own onto the layout, the way a picture one does: it reaches the file through the page's own shape tree. Whether it is still waiting for its content is asked of the media, since nothing marks such a placeholder empty. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I85707ff863a63874c6f493a0f284f25070f26e49 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9188 Tested-by: Caolán McNamara <caolan.mcnamara@collabora.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 44a05619224ab206dae4ef076269cbb8ce56f43b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/210009 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
47aa325014 |
tdf#166401: support custom placeholder prompt text
A slide layout is imported as a master page, and the shown text was only set
off a master, so an authored prompt was stored but never displayed or written
back; the export emitted Impress's own localized resource instead. Set it on
a master too, and accept the graphic kind in RestoreDefaultText, which keeps
the object an empty presentation object. Now that the two are
distinguishable, a placeholder that authored no prompt writes no text - its
text body keeps the properties a reader needs, and holds an empty paragraph.
A slide's own placeholder asks its master for that prompt before falling back
to the resource string, so the two no longer disagree about what is really
one placeholder.
This is what degraded in commit 73f1e76b1760d4e936aa22627623c922dcda964b
(cool#15754: Fix PPTX round-trip of empty picture placeholders, 2026-05-21):
mapping a layout <p:ph type="pic"/> to a real presentation service is what
made it reach that text reset, where the previous OutlinerShape mapping had
kept the authored prompt.
Implemented a TODO from commit
|
||
|
Mike Kaganski
|
898d74f652 |
cool#15754: do not paint a master's picture placeholder on slides
A slide carrying its own instance of a layout's picture placeholder showed two of them: its own movable one, and the layout's immovable one behind it. Hide a graphic placeholder on a master page the way the title, outline and notes ones already are. Everything else stays visible from the master, the footer and slide number above all. Regression from commit 73f1e76b1760d4e936aa22627623c922dcda964b (cool#15754: Fix PPTX round-trip of empty picture placeholders, 2026-05-21): a layout's <p:ph type="pic"/> used to become an outline shape, which was already hidden this way, so 25.04 shows one placeholder where we show two. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I0af73c8d2de8bc1e2686217bcb085d2ea8c7cc21 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9057 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com> (cherry picked from commit 15681ac99c23aa79ec90c0df833d6b091aaa56cf) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209952 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
b931a886ee |
cool#15754: keep content placeholders off the master
PowerPoint takes only title, body, dt, ftr and sldNum placeholders on a slide master and refuses to open a file whose master carries any other type, so a deck whose only layout held a picture placeholder saved into a file that could not be reopened at all. Skip such a placeholder in the master part and write the master page's own copy into the layout instead, which is where a slide resolves what it inherits. A master representing an equivalence group already writes its whole content into the layouts, so only the single-master case needed the copy. Write the slide placeholder's own geometry too, rather than leaving an empty <p:spPr/> for the layout to fill in. That is what PowerPoint itself authors, and our import does not always resolve the layout placeholder, which left the shape 0x0. Regression from commit 73f1e76b1760d4e936aa22627623c922dcda964b (cool#15754: Fix PPTX round-trip of empty picture placeholders, 2026-05-21), which mapped a pic placeholder to a real presentation service everywhere, the master included. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ie773152f0784de5104b5ab3ddcd0869621646f13 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9045 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Aron Budea <aron.budea@collabora.com> (cherry picked from commit 44f8c6457c797a9eae7e7612d377d227e1a354c3) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209951 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
45225d31c4 |
sw: keep the drag start point in a Point
The two coordinate statics were always written and read together, and every reader outside the edit window built a Point out of them anyway. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I7333728d5c51a15d435bd9ced2f1d8770f1e10c7 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8999 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit f7f785e565b208649065441ccbd0f0d056f3e29d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209829 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
db23488a24 |
sw: extract a helper for arming a frame drag
The same four lines stood at five press-handling sites; the last of them came with 3efae0655f0d26f0686560ae00ed57ca45f654a5 (sw: drag after text edit used a stale start point, 2026-08-02). Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I8f64d5855818f0d1e2166906033e3fb3eba0f104 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8995 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 16501dab9627e77dd3bcf640e091141b7aa7508a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209828 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
df813b5afc |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I4fd3eedda81455f082cdbcd07c323e8c4aab77ff Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9522 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit 90fca744e7cd75173ecd694bb326b7706bce9a6b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209523 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
71afdd2a6a |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I1a541a9cb977cde63b89b5e838fa0e77e3116c45 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9513 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 349f60f1c3f425bf3ec4a7b85e053c372a166e4a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209522 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
6c4d9d9e0c |
Avoid useless call to CreateColorBitmap
It is potentially expensive, and isn't needed: the ctor of Bitmap taking a bitmap and an apha mask only uses colors from source bitmap anyway, not its aplha. Change-Id: Icedae8c1e0c84e99654e223a36b52e9fbfdc6439 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9532 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 985f05348e44f8144e9f085f50ea372e31a916b5) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209521 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
1058d1f18d |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. And avoid useless call to CreateColorBitmap: it is expensive, and isn't needed: the ctor of Bitmap taking a bitmap and an apha mask only uses colors from source bitmap anyway, not its aplha. Change-Id: I08a9cc5087d5059ca250daed5bdfdd5039bd2836 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9534 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit 051a4de4d2af6fe143364fc367b070805865c285) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209520 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
164816bbd3 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I85badd996a26fe2f451da07fbd5dac4889a8e77d Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9531 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 4970121192f268115122b271b7b286d8691a82be) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209519 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
e035015902 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: Ia6284fbc0c36950ad40e3abdb9ea7095b4db894a Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9529 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit d9058f14a0d050adb9b594c0c3864af3c1a7583a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209518 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
f1e01466f0 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: If2caeff175ed120dde336c40f5e96ccd8157d5ad Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9528 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 86514450b4655f8e5623362c54d0153f5887d753) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209517 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
f03760a4b4 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: Id33122c14f009d43fff668c05efb662aa197a133 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9527 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit f8089f18c50639527f8657169c98b1f8c43c9787) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209516 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
1e10dbb206 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I4cd2e0d36e0b9fc68976b673d070578ca8ef628a Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9526 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 86c34245fb35609dba0054d96021f796a423236c) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209515 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
a244cac0cb |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I6ce4e568f0f3dd2baf79c4de45194822ace02268 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9524 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit e5ad1af72afd1975d2a75c4e995644bda7b02bf7) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209514 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
9bfa8c6843 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: Ieef19e0725dcdf0c9b17444b8f0e8b9eefaf74ee Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9521 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 0c72a1c0c7fd9da268a80038fad3ccb46a515834) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209513 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
c47c991c85 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I9323b8d5f5142657f402783438d67c388cf8d5e6 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9512 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit b6d5ade8b30112ff2b7bde1eb41008b43b713302) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209512 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
5df7faddb9 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I3fb2873a5e7b1e1e73266efdca1dc671a687277d Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9519 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Tested-by: Noel Grandin <noel.grandin@collabora.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit 74272c95e9ae155ec61c300e82a6bc1fd06939f4) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209511 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
7161748dcd |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. Change-Id: I899590dca6d00f7d5bf850d3ee0329215caaf2f6 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9517 Reviewed-by: Noel Grandin <noel.grandin@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 235b590828ebcc3167bb9b56f7acbf5c8300d4db) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209510 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
60c92f9752 |
CreateAlphaMask + CreateColorBitmap -> SplitIntoColorAndAlpha
See commit f7b67e6c8c5981790fb9ac162dcc68e470b53112 (pdfexport: reduce the number of times we split images, 2026-08-19) for rationale. While at that: drop the check added in commit |
||
|
Mike Kaganski
|
d385b81da3 |
Avoid second call to CreateAlphaMask
This is an expensive operation. Change-Id: I4ffd17de84d324dcb091905927d9ef7423fdb1c9 Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9505 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Noel Grandin <noel.grandin@collabora.com> (cherry picked from commit 9e2d6727f27caa7663a18564f46bb4a90b3b87a3) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209508 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
6214c82bd0 |
Fix signed/unsigned mismatch with MSVC 2026
After commit db92d0be756305ec0b1e122a1d00333477c17ba5 (dbase: apply
one consistent key-per-page limit in the .ndx index, 2026-08-19) a
local CODA-W build failed with:
C:/lo/online/engine/connectivity/source/drivers/dbase/dindexnode.cxx(518): error C2220: the following warning is treated as an error
C:/lo/online/engine/connectivity/source/drivers/dbase/dindexnode.cxx(518): warning C4018: '<': signed/unsigned mismatch
C:/lo/online/engine/connectivity/source/drivers/dbase/dindexnode.cxx(599): warning C4018: '<': signed/unsigned mismatch
The problem was the change of type of nNewCount from sal_uInt16 to
sal_uInt32, which made the following comparison warn:
if (nNewCount < (nMaxNodes_2 * 2))
nMaxNodes_2 is sal_uInt16, but multiplication gives int; before the
change, nNewCount was promoted from sal_uInt16 to int as well; but
now it isn't promoted, hence the problem.
Fix it by making sure that multiplication produces an unsigned value.
Change-Id: I056fa7517834b64706be27b4318bb21c42ecfebb
Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/9581
Tested-by: Jenkins CPCI <releng@collaboraoffice.com>
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
(cherry picked from commit d68d84af021b67ceaf9f94294722c2e97ee9e963)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209455
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
||
|
Mike Kaganski
|
45d479de0b |
cool#16072 sw: do not let a footnote force a row to split
lcl_GetMaxRowHeight returned the current height of the page body. That height is smaller when the page has footnotes. SwTabFrame::Split compares a row that must not split with it. When the row does not fit, Split treats the row as not fitting the whole page, and splits it so that no text is hidden. So a row of 4150 twips was compared with 3950, not with the 4535 that the body has when the page has no footnotes. It was split, 25 twips short of fitting. SetForceRowSplitAllowed is never reset, so the row stayed splittable and keep-together was ignored from then on. Measure the body as if the page had no footnotes: the print height of the page, without the header and the footer. The footnotes give that space back when the row needs it. A footnote of an earlier row stays on the page that the row leaves. A footnote of the row itself moves with the row. This is not enough on its own, because the row can now stay on a page that is too small for it. Split does not give the first row of a table to a follow: that would leave the table with no rows, and MakeAll uses Lower() right after it without a check. So the row stayed, and the text that did not fit was drawn below the table border, or was not drawn at all. Split now says what to do about it, and MakeAll does it. Which of the two ways it is depends on the footnote that took the space. A footnote of content on an earlier page cannot move, so the table has to: SwFootnoteRescue::MoveTable. The move needs bMoveAlways, because a table at the top of a page has no frame before it, and that normally stops the move. The page still has that footnote, so it is not left blank; and the next page has no such footnote, so the table does not move again. A footnote of the row itself can move, and then the row can stay: SwFootnoteRescue::MoveOwnFootnote. MakeAll asks the page to put the footnotes below the table. What has no room left there goes to the next page, and the body gets that space back. So the row keeps the page, like in Word. MakeAll asks only once, so a footnote that cannot move does not make a loop. This improves on commit |
||
|
Mike Kaganski
|
e55cb4a7b8 |
tdf#167714 sw: mark Word horizontal rules in the model
Nothing told a horizontal rule apart from an ordinary rectangle, so it could not be laid out the way Word does; the marker is that distinction, and its absence is what keeps documents written before it rendering unchanged. The UNO property reports DEFAULT_VALUE while false, since ODF export only writes direct values - else every shape and graphic style would carry it. Also let the RTF import build a rectangle for a rule, as the VML import already does, so both filters produce one model object. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Id2facd7bdc272197273462f0ad581537551234bc Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8657 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit dbe7dca2951640990d3febaf12a9c3118e15ec22) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209328 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
bf23464799 |
sw: drag after text edit used a stale start point
testSnapTextBoxToGrid in CppunitTest_sw_snap_to_grid failed in a CODA-W build with C:/lo/online/engine/sw/qa/extras/snaptogrid/snaptogrid.cxx(200) : error : Assertion Test name: `anonymous namespace'::testSnapTextBoxToGrid::TestBody equality assertion failed - Expected: 2032 - Actual : 4064 The press that ends a text edit also arms a frame drag, but left the drag start at the click that selected the object, so the drag jumped the object to that old point instead of following the press. Only a Kit build keeps the handles during a text edit, which is what made the jump a lost resize here. In Classic: - insert a rectangle with some text in it, and click outside it - double-click near its top left corner, to start editing its text - press on its bottom right corner, inside the shape but outside its text, and drag: the shape jumps, keeping its size, until the point of the double-click is under the pointer Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I9bce49be53969414995049c0d9e3aa8c7135413c Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8485 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit 3efae0655f0d26f0686560ae00ed57ca45f654a5) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209262 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
dc498d5a13 |
tdf#167713 sw: keep an RTF pict shape in its cell
A picture carrying shape properties is imported as a shape, and its insertion was not buffered like the rest of the row, so it landed in the row's first cell whichever cell it was written in. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ia5e51c49698d764561839e6361e6517648ca81fb Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8656 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 6c7c171d1468765f7e2a1872d7243f6bf809e37e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209261 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
e59e33f064 |
tdf#171370: a char must stringify to itself
Regression after commit
|
||
|
Mike Kaganski
|
44c92622d2 |
cool#16071 sw: do not lose footnote lines cut away to fit
RearrangeFootnotes may cause SetRstHeight to set the rest height to 0 for footnote lines no more fitting the footnote area; but 0 was used as a special value in IsInside, meaning "no rest height was set", which then incorrectly measured the upper instead - still as tall as the lines the frame had before the truncation. Then the frame was considered as having enough space, marked valid, and the lines cut away from it got no follow to move to, so they never appeared anywhere. Give the rest height its own "not given" state instead of spelling that as a number, and honour a height of zero for what it says. A magic negative would not do: SwContentFrame::WouldFit_ passes a negative height when the frames before it overfill the new upper, and that has to keep meaning that no line fits. The footnote computation in the constructor still falls back to the upper when it ends up at or below zero, so only the SetRstHeight path changes behaviour. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: If692577f3ee692ad5c0dbe87f9a11a2670f8cf16 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8906 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 77634c741bb538853a757c2a41e4c9b919f22815) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209271 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
50ce488d18 |
cool#16067 sw: take table rows back when the page regains space
A row moved to the follow because a footnote anchored in a row below it was taking space on the page; that row moved too and took the footnote along, so the page can now hold the row that was moved off it. The follow's rows now get one more try to join once the page has grown. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Id4120e3d35e13c8f2eb3eb7c2bb1deff3e13d0cb Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8868 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit c808931b72d1576b30be0c8ac813a4e48bc0e06e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209270 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
2f194066ca |
Related: tdf#171708 sw: keep a no-wrap footer object clear of the text
To see the problem, open that bug's document, https://bugs.documentfoundation.org/attachment.cgi?id=206628 - its footer holds a picture that wraps top and bottom, anchored to the footer paragraph and lifted by 81pt, which is less than its own height of 88.5pt. Writer applied that offset as given, so the picture covered the footer text and sat 7.3pt lower on the page than Word puts it. What Word does, measured over hand-authored probes: - an object that wraps top and bottom cannot share a line with its anchor paragraph, and a footer paragraph has nowhere to be pushed - the footer holds it at the page's footer distance; - so the object goes up instead, until its bottom reaches the top of that paragraph, and it does so for any offset of zero or less, no matter what the offset says; - an offset above zero puts the object below the paragraph and is left alone, and square or through wrapping is not touched at all. In the body the two already agree, because there the anchor paragraph can move and both push it below the object. So pull such an object up until it clears its anchor paragraph, for MSO-formatted documents only; ODF documents without compatibility flags keep their pre-existing layout. This reuses the compat flags as commit 409835939a3f1fbaeb1fdf705e60ce508a2be09e (tdf#162211 tdf#137590 layoutInCell: identify bMSOLayout better, 2026-04-14) did, identifying MSO layout by TAB_OVER_MARGIN and TAB_OVER_SPACING; no new compat flag. Headers are left alone: there Word clips the object at the page edge instead, which is a separate question, as is whether a footer object may hang off the bottom of the page at all (tdf#143933). Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ifabfe5ee28ba48f3bce615ff490eef52e2ab26f0 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8829 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 7926e7a5852178e17d682cf1c54e5b74d34e1abf) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209267 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> Tested-by: Jenkins |
||
|
Mike Kaganski
|
0b151e25c8 |
Related: tdf#171708 sw: keep floattable clear of a header/footer object
To see the problem, open that bug's document, https://bugs.documentfoundation.org/attachment.cgi?id=206628 - its body starts with a floating table spanning three pages (w:tblpPr, w:tblOverlap val="never"), and its footer holds a picture with <wp:wrapTopAndBottom/> and a negative positionV offset, so the picture reaches up into the body. Writer put every page's part of the table at the top of the body and painted its last rows over the picture, and it left the table lower on the page than Word leaves it. What Word does, measured from its print output over hand-authored probes with exact row heights: - how much of the table goes on a page comes from the full body height, with nothing in the way taken into account; - the part is then shifted up as a whole until its bottom clears the object below it, which may take it into the top margin area; - the page edge is the limit - where clearing the object would take the part off the page, it stays put and keeps the overlap. The low placement and the overlap are therefore one behaviour: the shift is what avoids the object. CalcOverlap already avoids overlaps for split flys, but a header/footer object is anchored outside the body and so was dismissed as a fly from another upper, and the one conflict it knows how to resolve is by moving down - which here would push the fly off the page. So consider such an object too, and resolve a conflict with one that starts below the fly by shifting the fly up until its bottom clears it, as far as the page frame allows. Where the page cannot hold the shifted fly, keep the overlap, and in no case give up a row to fit. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: I00eed5401b5981b0f3eb9b40e7be920bf71b15b3 Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8826 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit b356ac63ed2824e20906b26f337d7852994c1a6b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209266 Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
f0e78f28ee |
cool#16066 sw: omit the repeated heading where a row needs the full page
A row that may not break across pages, and is taller than what a page has
left below the repeated heading, had nowhere to go: it stayed under the
heading, reached past the bottom of the page, and everything below that was
not shown. The setup never worked - StarOffice 5.2 drops the same text.
Two strategies fix it and both are viable: split the row anyway, or leave
the heading off the page the row needs and repeat it again afterwards. This
implements the second, because Word does that. Either would serve our own
documents, but differing from Word turns every such DOCX into an
interoperability report, and adopting Word's layout later would then need a
compatibility flag to keep the ODF documents written in the meantime laying
out as their authors designed them.
A heading is not repeated where the following row could not fit under it, in
neither of the two places that create the repeats: splitting the table, and
rebuilding its frames after a row is added. Otherwise the row is only pushed
on again, and the repeat is left holding a page of its own.
Because the change creates layouts with a lone heading row and no non-heading
row on a page, GetFirstNonHeadlineRow() now returns nullptr where it
previously could not. The two places that dereferenced it unchecked,
MakeSelUnions and SwContentFrame::Cut, are guarded; the first of them crashed
on inserting a row into such a table.
A row taller than the page itself is still split regardless of the setting,
as since commit
|
||
|
Mike Kaganski
|
9b3f4c90ec |
tdf#171708: improve empty master detection
A split fly's intermediate anchor has no draw objects of its own - they all sit in the chain's master - so it was taken for a master that had already failed, and the last anchor was denied the split it needed. Signed-off-by: Mike Kaganski <mike.kaganski@collabora.com> Change-Id: Ic7cf989fcdbf013fa640e94877b7805358c3be8f Reviewed-on: https://gerrit.collaboraoffice.com/c/online/+/8821 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins CPCI <releng@collaboraoffice.com> (cherry picked from commit e8331d56cc14ddef0f243c9cc0091c5179457027) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/209265 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> |
||
|
Mike Kaganski
|
7152b64569 |
sw: fix null deref in InsertCnt_ when a section has no frame
A customer reported recurring crashes (no reproducer available) while a Writer
document was being saved by AutoRecovery. The minidump shows a read of address
0 on the GUI thread with this stack:
InsertCnt_ sw/source/core/layout/frmtool.cxx
MakeFrames sw/source/core/layout/frmtool.cxx
UnHideRedlines sw/source/core/layout/wsfrm.cxx
UnHide sw/source/core/layout/wsfrm.cxx
SwRootFrame::SetHideRedlines sw/source/core/layout/wsfrm.cxx
sw::DocumentRedlineManager::SetRedlineFlags sw/source/core/doc/DocumentRedlineManager.cxx
SwXMLWriter::WriteMedium sw/source/filter/xml/wrtxml.cxx
SwWriter::Write sw/source/filter/basflt/shellio.cxx
SwDocShell::SaveAs sw/source/uibase/app/docsh.cxx
SfxBaseModel::storeToRecoveryFile
Serializing to ODF temporarily shows the tracked changes, which rebuilds the
layout (SetHideRedlines -> UnHideRedlines -> MakeFrames -> InsertCnt_). While
closing a section, InsertCnt_ does
pLay = pLay->FindSctFrame();
and then dereferences pLay unconditionally (pLay->IsHiddenNow(), and in the other
branch pLay->ContainsContent()). FindSctFrame() returns nullptr when pLay is not
inside a section frame, which happens for a section whose content was not laid out
- seen here with tracked changes inside a table - so the dereference crashes.
Guard the result: when there is no section frame there is nothing to close, and
pLay must be left untouched because the following nodes reuse it.
The unguarded pLay->IsHiddenNow() was added in
|