sphobjinv: Manipulate and inspect Sphinx objects.inv files
Current Development Version:
Most Recent Stable Release:
Info:
Using Sphinx?
Having trouble writing cross-references?
sphobjinv (short for 'sphinx objects.inv') can help!
The syntax required for a functional Sphinx cross-reference is highly
non-obvious in many cases. Sometimes Sphinx can guess correctly what
you mean, but it's pretty hit-or-miss. The best approach is to provide
Sphinx with a completely specified cross-reference, and that's where
sphobjinv comes in.
After a pip install sphobjinv, find the documentation set you want
to cross-reference into, and pass it to sphobjinv suggest.
For internal cross-references, locate objects.inv within build/html:
$ sphobjinv suggest doc/build/html/objects.inv as_rst -st 58 Name Score --------------------------------------------------- ------- :py:property:`sphobjinv.data.SuperDataObj.as_rst` 60 :py:class:`sphobjinv.cli.parser.PrsConst` 59 :py:class:`sphobjinv.data.DataFields` 59 :py:class:`sphobjinv.data.DataObjBytes` 59 :py:class:`sphobjinv.data.DataObjStr` 59 :py:class:`sphobjinv.data.SuperDataObj` 59 :py:class:`sphobjinv.enum.HeaderFields` 59 :py:class:`sphobjinv.enum.SourceTypes` 59 :py:function:`sphobjinv.fileops.writebytes` 59 :py:function:`sphobjinv.fileops.writejson` 59 :py:class:`sphobjinv.inventory.Inventory` 59
The -s argument in the above shell command indicates to print the
fuzzywuzzy match score along with each search result, and -t 50
changes the reporting threshold for the match score.
For external references, just find the API documentation wherever it lives on the web,
and pass sphobjinv suggest a URL from within the documentation set
with the --url/-u flag. For example, say I need to know how to
cross-reference the linspace function from numpy (see
here):
$ sphobjinv suggest https://numpy.org/doc/1.19/reference/index.html linspace -su No inventory at provided URL. Attempting "https://numpy.org/doc/1.19/reference/index.html/objects.inv" ... Attempting "https://numpy.org/doc/1.19/reference/objects.inv" ... Attempting "https://numpy.org/doc/1.19/objects.inv" ... Remote inventory found. Name Score -------------------------------------------------------------- ------- :py:function:`numpy.linspace` 90 :py:method:`numpy.polynomial.chebyshev.Chebyshev.linspace` 90 :py:method:`numpy.polynomial.hermite.Hermite.linspace` 90 :py:method:`numpy.polynomial.hermite_e.HermiteE.linspace` 90 :py:method:`numpy.polynomial.laguerre.Laguerre.linspace` 90 :py:method:`numpy.polynomial.legendre.Legendre.linspace` 90 :py:method:`numpy.polynomial.polynomial.Polynomial.linspace` 90 :std:doc:`reference/generated/numpy.linspace` 90
NOTE that the results from sphobjinv suggest are printed using the longer
block directives, whereas cross-references must be composed using the
inline directives. Thus, the above linspace() function must be
cross-referenced as :func:`numpy.linspace`, not
:function:`numpy.linspace`.
Need to edit an inventory after it's created, or compose one from scratch?
sphobjinv can help with that, too.
objects.inv files can be decompressed to plaintext at the commandline:
$ sphobjinv convert plain -o doc/build/html/objects.inv doc/scratch/ Conversion completed. '...objects.inv' converted to '...objects.txt' (plain).
JSON output is supported (sphobjinv convert json ...), and
inventories can be re-compressed to the
partially-zlib-compressed form that intersphinx reads
(sphobjinv convert zlib ...).
Alternatively, sphobjinv exposes an API to enable automation of
inventory creation/modification:
>>> import sphobjinv as soi
>>> inv = soi.Inventory('doc/build/html/objects.inv')
>>> print(inv)
<Inventory (fname_zlib): sphobjinv v2.2, 205 objects>
>>> inv.project
'sphobjinv'
>>> inv.version
'2.2'
>>> inv.objects[0]
DataObjStr(name='sphobjinv.cli.core', domain='py', role='module', priority='0', uri='cli/implementation/core.html#module-$', dispname='-')
The API also enables straightforward re-export of an inventory,
for subsequent use with intersphinx cross-references.
See the docs
for more details.
Full documentation is hosted at Read The Docs.
Available on PyPI
(pip install sphobjinv).
Source on GitHub. Bug reports and feature requests are welcomed at the Issues page there.
Copyright (c) Brian Skinn 2016-2022
License: The MIT License. See LICENSE.txt for full license terms.