Sphinx objects.inv Inspection/Manipulation Tool

sphinx, sphinx-doc, inventory, manager, inspector, python
pip install sphobjinv==2.3.1


sphobjinv: Manipulate and inspect Sphinx objects.inv files

Current Development Version:

GitHub Workflow Status

Most Recent Stable Release:




Join the chat at https://gitter.im/sphobjinv/community https://pepy.tech/badge/sphobjinv/month

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 (or pipx 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


Cannot infer intersphinx_mapping from a local objects.inv.


Project: sphobjinv
Version: 2.3

220 objects in inventory.


11 results found at/above current threshold of 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/1.23/reference/index.html linspace -su

Attempting https://numpy.org/doc/1.23/reference/index.html ...
  ... no recognized inventory.
Attempting "https://numpy.org/doc/1.23/reference/index.html/objects.inv" ...
  ... HTTP error: 404 Not Found.
Attempting "https://numpy.org/doc/1.23/reference/objects.inv" ...
  ... HTTP error: 404 Not Found.
Attempting "https://numpy.org/doc/1.23/objects.inv" ...
  ... inventory found.


The intersphinx_mapping for this docset is LIKELY:

  (https://numpy.org/doc/1.23/, None)


Project: NumPy
Version: 1.23

8074 objects in inventory.


8 results found at/above current threshold of 75.

  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 v2.3, 220 objects>
>>> inv.project
>>> inv.version
>>> inv.objects[0]
DataObjStr(name='sphobjinv.cli.convert', domain='py', role='module', priority='0', uri='cli/implementation/convert.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

The sphobjinv documentation (including docstrings and README) is licensed under a Creative Commons Attribution 4.0 International License (CC-BY). The sphobjinv codebase is released under the MIT License. See LICENSE.txt for full license terms.