dvisvgm News

dvisvgm 3.6.1 has been released

Tue, 04 Aug 2026 09:49:41 +0200

Fixed the Unicode mapping of some ligatures.
Fixed several potential memory issues.
Changed some parts of the code to be compatible with C++ standard >= 20.

dvisvgm 3.6 has been released

Mon, 08 Dec 2025 15:26:51 +0100

Removed the restriction to process PDF-1.x files only (issue #291).
Option --page: added evaluation of %P which expands to the maximum page number. Also, %(expr) is supported to evaluate arithmetic expressions, like %(P-1). See the manual page for further information.
Option --page: page numbers greater than the maximum page number are no longer replaced by the maximum page number but ignored.
PDF handler: the computation of opacity values has been extended (issue #292).
Fixed the lookup of character code 0 in character maps (issue #294).
Fixed the height and depth values shown when processing data created by the preview package.
Updated the bundled brotli library to version 1.2.0.
Some code improvements.

dvisvgm 3.5 has been released

Fri, 23 May 2025 09:57:33 +0200

Fonts are now embedded in deterministic rather than random order (issue #288).
Different glyphs now always get distinct Unicode code points to avoid collisions, even if they represent the same character (issue #287).
Fixed font-related segmentation faults caused by malformed DVI files (issue #289).
DVI error messages now also mention the byte position where the error occurred in the DVI file.
The evaluation of PostScript operators has been improved and should be more stable.
Many code refactorings and improvements.

dvisvgm 3.4.4 has been released

Thu, 06 Mar 2025 11:08:05 +0100

Extended the computation of Unicode points by looking for known glyph names.
Added some common glyph names not yet covered by the AGL table.
Added a warning shown if unsupported PDF specials have been ignored.
Some additions to the manual page.
Several code improvements.

dvisvgm 3.4.3 has been released

Sat, 04 Jan 2025 13:04:24 +0100

Fixed an index error in PS operator colorimage (issue #279).
Updated the bundled xxHash library to version 0.8.3.
Minor changes to the manual page.

dvisvgm 3.4.2 has been released

Thu, 14 Nov 2024 16:57:15 +0100

Fixed the evaluation of Unicode data performed when converting PDF files (issue #276).
Fixed the horizontal character positioning in the PDF handler to prevent exceeding text lines (issue #276).
Fixed the drawing of single-colored tensor product patches to also get correct shapes if parts of the surface overlap.
TTF/WOFF data written to SVG files now use a static creation and modification date to ensure identical SVG files for unchanged input files (issue #120).
Slightly improved formatting and wording of the manual page.

dvisvgm 3.4.1 has been released

Mon, 16 Sep 2024 12:13:14 +0200

This is just a small maintenance release with these changes:

When converting PDF files, dvisvgm now shows the number of the page being processed.
Added missing #include statement required as of GCC 15 (issue #273).

dvisvgm 3.4 has been released

Wed, 24 Jul 2024 11:56:33 +0200

Option --embed-bitmaps is now also applied to images included with special dvisvgm:img (issue #229).
Added macro {?cmyk(c,m,y,k)} to allow for directly specifying CMYK colors in literal SVG fragments.
Extended the dvips color specials to distinguish between fill and stroke colors. For further information see the manual page.
Added color special color set to replace the current color without pushing a new value onto the color stack. See the manual page for more details.
Added macros {?fillcolor} and {?strokecolor} to retrieve the current fill and stroke color, respectively. The former macro {?color} equals {?fillcolor} and is still available too.
Replaced millimeter units with big point (bp) units in the message showing the extent of the generated SVG file. TeX’s “big points” equal SVG, PostScript, and PDF points (1in = 72bp), while TeX points are slightly smaller (1in = 72.27pt).
```
$ dvisvgm -d2 frktest
pre-processing DVI file (format version 2)
  processing page 1
  graphic size: 398.34pt x 456.34pt (396.85bp x 454.63bp)
  output written to frktest.svg
1 of 1 page converted in 0.0754669 seconds
```
Fixed the detection of MIME types depending of file name suffixes.
Several small code improvements.

dvisvgm 3.3 has been released

Wed, 10 Apr 2024 15:59:07 +0200

Added file dvisvgm.map to the default font map files being looked up automatically. It can be used to add font mappings which should always be taken into account when dvisvgm is called.
Added evaluation of #include and #includefirst present in font map files to allow for including the contents of other map files.
The conversion of CMYK colors to RGB has been modified so that the resulting colors are now close to those created by Ghostscript and most PDF viewers. Previously, a simple mapping of the color components, like cmyk(1,0,1,0)=rgb(0,1,0) for green, was used. Ghostscript, in contrast, applies ICC profiles to display colors as produced by a designated printer.
Here’s an example showing the results of the old (left) and the new CMYK to RGB color conversion applied to a Coons shading patch:

Fixed the initialization of a transformation matrix which led to invalid bounding boxes when converting a sequence of DVI pages (issue #268).
Fixed the computation of internal Coons tensor points.
Updated the bundled GoogleTest to version 1.12.1.

dvisvgm 3.2.2 has been released

Thu, 07 Mar 2024 12:37:55 +0100

Due to the removal of some Ghostscript-specific operators from GS version 10.02.1, dvisvgm fails to extract bitmaps embedded in PS/EPS files when using the latest Ghostscript releases (issue #264).
This dvisvgm release replaces the involved operators and should fix the issue.

dvisvgm 3.2.1 has been released

Thu, 29 Feb 2024 09:57:09 +0100

The Ghostscript library is now only loaded when the first PostScript special is processed. This slightly speeds up the conversion of DVI files that don’t contain PS specials (issue #262).
Applied some changes required for TeX Live builds (patches by Karl Berry, issue #259, issue #260).
Added more details to some DVI warning messages.
Added some further information to the manual page.

dvisvgm 3.2 has been released

Thu, 11 Jan 2024 10:13:14 +0100

Fixed a rounding issue that could lead to slightly bigger text extents (issue #256).
Added the macros {?pageno}, {?svgfile}, and {?svgpath} to the dvisvgm special commands.
Added a separate verbosity flag 8 for user messages printed by special dvisvgm:message.
Added the command-line option --message to allow for printing user messages after writing an SVG file (issue #258).
Applied this upstream patch regarding undefined type punning to the bundled woff2 library.
Few improvements of the manual page.

dvisvgm 3.1.2 has been released

Thu, 12 Oct 2023 15:19:57 +0200

Improved the evaluation of graphic extents created by the preview package (issue #246).
Fixed the computation of the character box extents performed when using option --exact-bbox (issue #248).
Fixed a memory error in the optimizer module (issue #250).
Updated the bundled Brotli library to version 1.1.0.

dvisvgm 3.1.1 has been released

Tue, 29 Aug 2023 12:44:48 +0200

Added evaluation of optional dash prefixes to command-line option --optimize to exclude selected optimizer modules.
Fixed the invalid removal of clipping paths referenced in the defs section.
Fixed the automatic calls of optimizer module remove-clippaths.
Suppress warning messages about deprecated option -dNEWPDF introduced with Ghostscript 10.02.0 (issue #245).
Updated the bundled xxHash library to version 0.8.2.
Fixed a few paragraphs in the manual page.

dvisvgm 3.1 has been released

Thu, 06 Jul 2023 18:34:24 +0200

Added the command-line option --currentcolor to replace a given color with CSS variable currentColor (issue #214).
Added special dvisvgm:currentcolor to replace the currently active color with CSS variable currentColor (issue #214).
Added special dvisvgm:message to print given texts to the console.
Added support for Omega Font Metric (OFM) files.
Added a fallback mechanism for missing characters in OFM-based virtual fonts proposed by Takuji Tanaka (issue #241).
Fixed the handling of whitespace characters in PDF file paths (issue #233).
Fixed the vertical placement of “vertical fonts” in horizontal mode (issue #235).
Fixed the parsing of floating point numbers which now prevents errors when using libc++ instead of libstdc++ (issue #240).
Fixed potential segmentation faults triggered by missing font files (issue #241).
Fixed a GLIBCXX assertion error occurred if dvisvgm is built with _GLIBCXX_ASSERTIONS enabled (issue #242).
Improved the handling of font subset prefixes in the new, mutool-based PDF handler (issue #242).
Several minor code improvements.

dvisvgm 3.0.4 has been released

Fri, 10 Mar 2023 10:41:33 +0100

Fixed a bug that could prevent reading the output of subprocesses (issue #232).
Removed the optional dependency on libcrypto (OpenSSL) due to the deprecation of its MD5 functions. Instead, the bundled MD5 function is always used now.
Added TeX Live conditionals to the build system (patch by Karl Berry).
Small code and manpage improvements.

dvisvgm 3.0.3 has been released

Tue, 21 Feb 2023 12:20:06 +0100

This is a small bugfix release addressing these issues:

Fixed resetting the character context that could lead to invalid character positions and properties (issue #227).
Added missing #include directives required for GCC 13.

dvisvgm 3.0.2 has been released

Mon, 30 Jan 2023 09:15:14 +0100

This is a small maintenance release that addresses the following issues:

Fixed the computation of font sizes performed by the new PDF handler (issue #220).
Fixed the viewBox coordinates determined by the new PDF handler (issue #223).
Improved the detection of font names used in PDF files (issue #224).
Minor code improvements, increased compatibility with older GCC versions.

dvisvgm 3.0.1 has been released

Tue, 10 Jan 2023 10:49:43 +0100

This is the first stable release of dvisvgm 3. An overview of the main changes and additions can be found in the announcement of pre-release version 3.0 below. Version 3.0.1 fixes several issues reported in the last weeks.

Added support for the 64-bit Ghostscript library in TeX Live (patch sent by Akira Kakuto).
MacTeX: explicitly lookup libgs in /usr/local/lib (issue #197).
Fixed the character spacing when processing XDV glyph arrays. (issue #198).
Fixed the invalid removal of non-redundant clipPath elements in some cases if option --optimize is present (issue #203).
Fixed the computation of the graphics extents printed to the console when using the preview package (issue #209).
Several improvements of the autotools scripts (thanks to Sam James).
Improved the processing of fonts performed by the new PDF handler.

dvisvgm 3.0: Pre-release of new major version

Mon, 17 Oct 2022 09:48:19 +0200

dvisvgm 3.0 is a new major version of the utility that’s still considered beta status. While the new version doesn’t bring many obvious new features, essential core functionality has changed under the hood. The following list gives an overview of the most important changes and additions.

The previously bundled reduced FontForge library has been replaced with a dedicated font writer library used to create TrueType and WOFF files from the font data processed by dvisvgm.
A new PDF handler based on mutool has been added to retain dvisvgm’s PDF conversion functionality after the removal of Ghostscript’s old PDF interpreter announced for version 10.1.0.
- The new handler supports all font options of dvisvgm, i.e. in contrast to the current Ghostscript-based PDF processor, fonts can not only be embedded as paths but also as SVG, TTF, WOFF, and WOFF2 fonts (depending on options --no-fonts and --font-format).
- The new handler is still experimental and may contain bugs. Please report any issues via the issue tracker at GitHub.
- Not all PDF features are supported – either because there is no simple equivalent in SVG or because mutool doesn’t provide means to access the corresponding data. For example, Type 3 fonts and color gradients fills are not yet processed by dvisvgm’s new PDF handler.
- The former Ghostscript-based PDF handler is still available. It’s automatically invoked if dvisvgm finds a proper Ghostscript version. Otherwise, it looks for the mutool executable through the system’s search path and uses the new handler in case it’s found. The desired PDF handler can also be selected by setting environment variable DVISVGM_PDF_PROC to either gs or mutool.
Option --embed-bitmaps has been added to allow for embedding of external images referenced in DVI files into the generated SVG files. By default the SVG only contains a reference to the image.
The optional filters :even and :odd have been added to option --page which limit the given page ranges to even and odd page numbers respectively.

You can download the sources and pre-built binaries for MiKTeX from the project page at GitHub.