<TR><TD ALIGN=CENTER><A NAME=a:pdf HREF=#d:pdf>pdf</A>
</TD><TD>Portable Document Format (PDF)</TD> </TR>
<TR><TD ALIGN=CENTER><A NAME=a:pic HREF=#d:pic>pic</A>
-</TD><TD>Kernighan's PIC graphics language</TD> </TR>
+</TD><TD>Kernighan's PIC graphics language </TD> </TR>
<TR><TD ALIGN=CENTER><A NAME=a:plain HREF=#d:plain>plain</A>
<BR><A NAME=a:plain-ext HREF=#d:plain-ext>plain-ext</A>
</TD><TD>Simple text format</TD> </TR>
# ps_to_png.sh and other make rules rely on netpbm tools
# and psconvert, which is part of the GMT library.
#
-# python is used to run jconvert.py, which converts the json schema
-# graphviz_json_schema.json to html. This also relies on the python
-# package json2html.
+# Install python modules with:
+# $ pip install --user -r requirements.txt
+#
+# python is used to run:
+# - jconvert.py, which converts the json schema graphviz_json_schema.json to html.
+# This also relies on the python package json2html.
+# - mkoutput.py, which converts templates/output.html to output.html
+# This requires the jinja2 python package.
#
# The main product are 7 web pages:
# arrows.html - arrow_grammar
cat colors.n >> colors.html
rm -rf colortmp
-schema.html : jconvert.py graphviz_json_schema.json
- ./jconvert.py graphviz_json_schema.json schema.html
-output.html : output.1 output.2 outputs mkoutput.sh plugins.png jconvert.py schema.html
+output.html : output.1 output.2 outputs mkoutput.py plugins.png jconvert.py schema.html templates/output.html.j2
cat output.1 > output.html
- ./mkoutput.sh < outputs >> output.html
+ ./mkoutput.py < outputs >> output.html
cat output.2 >> output.html
+schema.html : jconvert.py graphviz_json_schema.json
+ ./jconvert.py graphviz_json_schema.json schema.html
+
html.html : html.1 html_grammar html.2 html.3 html1.gif html2.gif html3.gif html4.gif mklang
./mklang html_grammar gramtmp
cat html.1 > html.html
(for s in $$(cat shapelist); do rm -f $$s.gif; done)
EXTRA_DIST = $(XGIF) mklang.y mkarrows.sh mkattrs.sh mkshapes.sh mkstyles.sh mktapers.sh \
- mktypes.sh mkarrowtbl.sh mkoutput.sh mkshhtml.sh \
+ mktypes.sh mkarrowtbl.sh mkoutput.py mkshhtml.sh \
ps_to_png.sh arrow_grammar grammar html_grammar \
shapelist attrs.1 colors.1 colors.n \
output.1 output.2 html.1 html.2 html1.dot html.3 \
--- /dev/null
+#!/usr/bin/env python3
+# Takes `outputs` as stdin and generates output.html
+# Uses `templates/output.html.j2`
+# See `outputs` file for format documentation.
+
+import jinja2
+import markupsafe
+import re
+import sys
+from typing import Dict, Tuple
+
+HEADER_RE = re.compile(r'^:(?P<params>[^:]+):(?P<format>.*)')
+
+# Tuple of command-line-params for an output format, e.g. ('jpg', 'jpeg', 'jpe')
+params : Tuple[str, ...] = ()
+
+# Map from tuple of command-line-params to full name of the output format
+formats : Dict[Tuple[str, ...], str] = {}
+
+# Map from tuple of command-line-params to an HTML description string
+html_descriptions : Dict[Tuple[str, ...], str] = {}
+
+for line in sys.stdin:
+ # Skip comment lines.
+ if line.startswith('#'):
+ continue
+
+ m = HEADER_RE.match(line)
+ if m:
+ # This is a header line. Grab out the values.
+
+ # Command-line formats are slash-separated.
+ params = tuple(m.group('params').split('/'))
+
+ # Full format name is plain text
+ formats[params] = m.group('format')
+
+ # Set an empty string html description, ready to append to.
+ html_descriptions[params] = ''
+ else:
+ # This is an HTML line, possibly a continuation of a previous HTML line.
+ html_descriptions[params] += line
+
+
+env = jinja2.Environment(
+ # Load template files from ./templates/
+ loader=jinja2.FileSystemLoader('templates'),
+ # Auto-HTML-escape any html or xml files.
+ autoescape=jinja2.select_autoescape(['html', 'xml', 'html.j2', 'xml.j2']),
+ # Whitespace control
+ trim_blocks=True,
+ lstrip_blocks=True,
+ # Raise exception on any attempt to access undefined variables.
+ undefined=jinja2.StrictUndefined,
+)
+template = env.get_template('output.html.j2')
+print(template.render(
+ formats=formats,
+ # Vouch for the HTML descriptions as being safe and not needing auto-HTML-escaping.
+ # This is reasonable because the HTML descriptions are not attacker-controlled.
+ descriptions={
+ params: markupsafe.Markup(desc)
+ for params, desc in html_descriptions.items()
+ }
+))
+++ /dev/null
-#! /bin/ksh
-typeset -A desc fullname
-name=
-
-# Given the output specifications as stdin, read each one,
-# storing the formats long name and description in the fullname
-# and desc arrays, respectively, indexed by the format name.
-# The initial line of an item has the format :name:long name
-#
-# Lines beginning with '#' are treated as comments.
-OLDIFS="$IFS"
-IFS=
-while read line
-do
- c=${line:0:1}
- if [[ $c == '#' ]]
- then
- continue
- elif [[ $c == ':' ]]
- then
- if [[ -n "$name" ]]
- then
- desc[$name]=$txt
- fullname[$name]=$tag
- txt=""
- fi
- line=${line#:}
- if [[ "$line" == *:* ]]
- then
- name=${line%%:*}
- tag=${line#$name:}
- else
- name=$line
- tag=""
- fi
- else
- txt="$txt${line}\n"
- fi
-done
-IFS="$OLDIFS"
-
-if [[ -n "$name" ]]
-then
- desc[$name]=$txt
- fullname[$name]=$tag
- txt=""
-fi
-
-#print ${!fullname[@]}
-#print ${desc[@]}
-#exit
-
-set -s ${!desc[@]}
-
-# Output a brief description of the formats as a table.
-# This is followed by a list of the formats, with the long
-# description of each.
-# The formats are alphabetized on output.
-# Note that an item may have multiple names, i.e., the first
-# field may have the format name1/name2/name3.
-# The output format <name> is given the anchor a:<name> in the
-# table and d:<name> in the list.
-
-print "<TABLE ALIGN=CENTER>"
-print "<TR><TH>Command-line<BR>parameter</TH><TH>Format</TH></TR>"
-for i
-do
- print -n " <TR>";
- print -n "<TD ALIGN=CENTER>";
- first=yes
- for n in ${i//\// }
- do
- if [[ -n $first ]]
- then
- first=
- else
- print -n "<BR>"
- fi
- printf "<A NAME=a:%s HREF=#d:%s>%s</A>\n" $n $n $n
- done
- print -n "</TD><TD>"
- print -n ${fullname[$i]}; print "</TD> </TR>";
-done
-print "</TABLE>"
-
-print "<HR>"
-
-#set -s ${!desc[@]}
-print "<H2>Format Descriptions</H2>\n<DL>"
-for i
-do
- first=yes
- for n in ${i//\// }
- do
- if [[ -n $first ]]
- then
- first=
- else
- print -n ","
- fi
- printf "<DT><A NAME=d:%s HREF=#a:%s><STRONG>%s</STRONG></A>\n" $n $n $n
- done
- print "<DD>${desc[$i]}"
-done
-print "</DL>\n<HR>"
-
-exit 0
# List of Graphviz output formats
-# Each item consists of line of the form :<name>:<long description>
+# Each item consists of line of the form :<param1>[/<param2>...]:<format>
# followed by a description of the format in HTML
-# The name may consist of multiple, related names, separated by '/'.
-# The long description is just text.
+# The command-line params may consist of multiple, related params, separated by '/'.
+# The format is just text.
# The items are alphabetized when the page is created.
#
:eps:Encapsulated PostScript
ease of parsing. In general, the <A HREF=#d:dot>dot</A> and
<A HREF=#d:xdot>xdot</A> are preferable in terms of the quantity of
information provided.
-:bmp: Windows Bitmap Format
+:bmp:Windows Bitmap Format
Outputs images in the Windows <A HREF="http://en.wikipedia.org/wiki/Bitmap">BMP</A> format.
-:ico: Icon Image File Format
+:ico:Icon Image File Format
Outputs images in the Windows <A HREF="http://en.wikipedia.org/wiki/ICO_(icon_image_file_format)">ICO format</A>.
-:pdf: Portable Document Format (PDF)
+:pdf:Portable Document Format (PDF)
Produces <A HREF="http://www.adobe.com/devnet/pdf/">PDF</A> output.
(This option assumes Graphviz includes the Cairo renderer.)
Alternatively, one can use the <A HREF="#d:ps2">ps2</A> option to
produce PDF-compatible PostScript, and then use a ps-to-pdf converter.
-:tif/tiff: TIFF (Tag Image File Format)
+:tif/tiff:TIFF (Tag Image File Format)
Produces <A HREF="http://www.libtiff.org/">TIFF</A> output.
-:vml/vmlz: Vector Markup Language (VML)
+:vml/vmlz:Vector Markup Language (VML)
Produces <A HREF="http://www.w3.org/TR/NOTE-VML">VML</A> output,
the latter in compressed format.
<P>
See <A HREF=#ID>Note</A>.
-:gtk: GTK canvas
+:gtk:GTK canvas
Creates a <A HREF="http://www.gtk.org/">GTK</A> window and displays the output there.
-:webp: Image format for the Web
+:webp:Image format for the Web
Produces output in the image format for the Web (WEBP) format, optimized for
web devices such as tablets.
See Wikipedia's <a href="http://en.wikipedia.org/wiki/Webp">WebP</a>
or Google's <a href="http://code.google.com/speed/webp/">webp</a> pages.
-:xlib/x11: Xlib canvas
+:xlib/x11:Xlib canvas
Creates an <A HREF="http://en.wikipedia.org/wiki/Xlib">Xlib</A> window and displays the output there.
--- /dev/null
+json2html==1.3.0
+Jinja2==2.10
+MarkupSafe==1.0
--- /dev/null
+<TABLE ALIGN=CENTER>
+<TR><TH>Command-line<BR>parameter</TH><TH>Format</TH></TR>
+{% for params, format in formats | dictsort %}
+ <TR><TD ALIGN=CENTER>
+ {%- for p in params -%}
+ <A NAME=a:{{p}} HREF=#d:{{p}}>{{p}}</A>
+ {%- if not loop.last %}
+
+<BR>
+ {%- endif %}
+ {%- endfor %}
+
+</TD><TD>{{ format }}</TD> </TR>
+{% endfor %}
+</TABLE>
+<HR>
+<H2>Format Descriptions</H2>
+<DL>
+{% for params, description in descriptions | dictsort %}
+ {% for p in params %}
+ {% if not loop.first -%}
+ ,
+ {%- endif -%}
+ <DT><A NAME=d:{{p}} HREF=#a:{{p}}><STRONG>{{p}}</STRONG></A>
+ {% endfor -%}
+ <DD>
+ {{- description }}
+{% endfor %}
+</DL>
+<HR>
projects / graphviz / commitdiff