| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| | |
|
| |
|
|
| |
In case the latex process fails for formulas there is no reason to try to start a dvips process.
|
| |\
| |
| | |
issue #7753: SVG formula all with zero width
|
| | |
| |
| |
| |
| | |
The wrong index to determine the old size was used. (index for `#` instead of `=`).
Problem was not just for SVG but also for PNG.
|
| | | |
|
| | |
| |
| |
| | |
On subsequent calls we should know which inkscape version we have when this is not the case we already tried to determine it, but it failed so we should bail out here without error as an error has already been given.
|
| |/
|
|
| |
The command line interface (CLI) of Inkscape 1.0 has changed in comparison to previous versions. In order to invoke Inkscape, the used version is detected and based on the version the right syntax of the CLI is chosen.
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
When we have set `HTML_FORMULA_FORMAT = svg` we still get the message:
```
Generating image form_0.png for formula
```
instead of
```
Generating image form_0.svg for formula
```
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
Based on the implementation as mentioned in #7578 to add the possibilities to generated svg images for formulas
- When running with inkscape 92.3 / 92.4 on Cygwin /Windows there were no images generated as the `-o` flag didn't exist, the output file had to be specified as part of the `-l` (or `--export-plain-svg`) option
- For more flexibility the system is checked on existence of the `pdf2svg` and `inkscape` executables, so no compilations flags are necessary
|
| |
|
|
| |
Added development option to use inkscape to create SVG formulas.
|
| |
|
|
| |
Also added option HTML_FORMULA_FORMAT to generate SVG files for images (requires pdf2svg)
|
| |
|
|
|
|
|
|
|
| |
The command `ps2epsi` also uses the `gs` command and this gives some problems with the gs9.50 version.
An alternative is the `eps2eps` command that doesn't have these problems with the gs9.50 version.
See also:
- #7290
- https://bugs.ghostscript.com/show_bug.cgi?id=702024
|
| | |
|
| |\
| |
| |
| | |
https://github.com/albert-github/doxygen into albert-github-feature/bug_formula
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
When having a problem like:
```
/** \file */
/**
* The fie
*
* @form: the new format
* @content: the content of the form
*
* Sets @form content to be @content.
*
* \f$ \alpha \f$ \f[ \beta \f]
*/
void fie(char *form, char *content);
```
we get warnings like:
```
.../aa.h:5: warning: Wrong formula id -1
.../aa.h:6: warning: Found unknown command '\content'
.../aa.h:8: warning: Wrong formula id -1
.../aa.h:8: warning: Found unknown command '\content'
```
The `@form:` is in RST analogue to the doxygen `@param form` but not understood by doxygen.
The result is that it is seen as an formula as doxygen translates the formula commands to `\form`.
For `@content` is handled properly as this is not a doxygen command.
In this patch a more unique name is chosen so these type of conflicts don't appear (that easy). The resulting, better, warnings are now:
```
.../aa.h:5: warning: Found unknown command '\form'
.../aa.h:6: warning: Found unknown command '\content'
.../aa.h:8: warning: Found unknown command '\form'
.../aa.h:8: warning: Found unknown command '\content'
```
|
| |\ \
| | |
| | | |
Create possibility to define LaTeX commands for formulas
|
| | | |
| | |
| | |
| | |
| | |
| | | |
To be able to have building bocks for formulas one can create a `\newcommand` (or when one wants to change a command `\renewcommand`).
Due to the different handling of LaTeX commands in pure LaTeX code (latex output and formulas converted to images) and MathJax it is necessary to transform LaTeX commands to the MathJax equivalent.
This is done in a transparent way by providing the new commands in a file and add this verbatim to the pure LaTeX code and to translate the `\newcommand` and `\renewcomamnd` to MathJax macros.
|
| | | | |
|
| |\ \ \
| | | |
| | | |
| | | | |
https://github.com/albert-github/doxygen into albert-github-feature/bug_terminate
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
In case a `WARN_LOGFILE` is used it in this file no clear whether doxygen finished correctly or exited beforehand in case a fatal error.
Now a 'Exiting...' is also given in the warning log file in case an error is a fatal error.
|
| | | | | |
|
| |/ / /
| | |
| | |
| | |
| | |
| | | |
namespace
Fixing some memory issues caused by not correctly freed pointers
|
| | |/
|/|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
-sDEVICE=ppmraw -sOutputFile=_form0.pnm -dNOPAUSE -dBATCH -- _form0.ps. Check your installation!
@maehr had a talk with Robin Watts and Ken Sharp at IRC and there seem to be basically 3 different problems:
* `-r%dx%d` (the dimension for `r` shouldn't be `-r384x384x`, but `-r384x384`),
* misuse / unnecessary use of `--` and
* since 9.50 the command needs more control access (that might be worked around by either whitelisting the file via `--permit-file-read=_form0.eps` (only works from 9.50 and upwards) or generally accepting any file with `-dNOSAFER` (works since quite some time). The second option is considered to be unsafe if we would process any file, but in this case we process self produced / controlled files. I don't know if doxygen has any threat model that it assumes. ).
> Ken Sharp: Yeah the %dx is wrong, as Robin says its sheer luck that works
the -- isn't needed and is what's causing the first problem
and file control is the new bugbear
The suggestions have been implemented and test / docs works now with old and new version.
|
| |/
|
|
|
|
| |
`divips -E` does not always return the correct bounding box. The program `ps2epsi` returns a good bounding box.
- don't use the `-E` with `dvips`
- convert the file from `dvips` with `ps2epsi` to an encapsulated postscript file
|
| |
|
|
|
| |
Some older distributions with newunicode package 1.1 need this option `\usepackage[utf8]{inputenc}`.
Option was already present in the main LaTeX output i.e. refman.tex
|
| |\
| |
| | |
Small problems when having code in LaTeX output
|
| | |
| |
| |
| |
| |
| |
| |
| | |
- in formulas it is also possible to have special characters
- adding special character '-' for U+2212
- making special characters more system independent (plain latex was not handled) and should be independent of hyperlink setting
- without hyperlinks the name of normally linked names with underscore kept e.g. their underscores instead of escaping it
- be sure that on the end of a code section the previous line is properly 'closed'.
|
| | |
| |
| |
| |
| |
| | |
Regression on pull request #698.
We shouldn't switch to pdflatex for formula generation (follow up procedure: dvips and gs / splitting into images does not work).
In the travis tests this was shown as well but not flagged as an error.
|
| |/
|
|
| |
For a number of languages the output in LaTeX has been improved so the regular documentation can be generated (automatically) in LaTeX / PDF as well.
|
| |
|
|
| |
formula.cpp handles package arguments correctly
|
| |
|
|
| |
improve performance
|
| |
|
|
| |
Reset current directory in case of return.
|
| |
|
|
| |
Signed-off-by: Adrian Negreanu <adrian.m.negreanu@intel.com>
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
