Current LanguageTool version: 5.5
This is a Python wrapper for LanguageTool. LanguageTool is open-source grammar tool, also known as the spellchecker for OpenOffice. This library allows you to make to detect grammar errors and spelling mistakes through a Python script or through a command-line interface.
By default, language_tool_python will download a LanguageTool server .jar and run that in the background to detect grammar errors locally. However, LanguageTool also offers a Public HTTP Proofreading API that is supported as well. Follow the link for rate limiting details. (Running locally won't have the same restrictions.)
Local server is the default setting. To use this, just initialize a LanguageTool object:
import language_tool_python
tool = language_tool_python.LanguageTool('en-US') # use a local server (automatically set up), language EnglishThere is also a built-in class for querying LanguageTool's public servers. Initialize it like this:
import language_tool_python
tool = language_tool_python.LanguageToolPublicAPI('es') # use the public API, language SpanishFinally, you're able to pass in your own remote server as an argument to the LanguageTool class:
import language_tool_python
tool = language_tool_python.LanguageTool('ca-ES', remote_server='https://language-tool-api.mywebsite.net') # use a remote server API, language CatalanIf you want to decide which Match objects to apply to your text, use tool.check (to generate the list of matches) in conjunction with language_tool_python.utils.correct (to apply the list of matches to text). Here is an example of generating, filtering, and applying a list of matches. In this case, spell-checking suggestions for uppercase words are ignored:
>>> s = "Department of medicine Colombia University closed on August 1 Milinda Samuelli"
>>> is_bad_rule = lambda rule: rule.message == 'Possible spelling mistake found.' and len(rule.replacements) and rule.replacements[0][0].isupper()
>>> import language_tool_python
>>> tool = language_tool_python.LanguageTool('en-US')
>>> matches = tool.check(s)
>>> matches = [rule for rule in matches if not is_bad_rule(rule)]
>>> language_tool_python.utils.correct(s, matches)
'Department of medicine Colombia University closed on August 1 Melinda Sam'From the interpreter:
>>> import language_tool_python
>>> tool = language_tool_python.LanguageTool('en-US')
>>> text = 'A sentence with a error in the Hitchhiker’s Guide tot he Galaxy'
>>> matches = tool.check(text)
>>> len(matches)
2
...
>>> tool.close() # Call `close()` to shut off the server when you're done.Check out some Match object attributes:
>>> matches[0].ruleId, matches[0].replacements # ('EN_A_VS_AN', ['an'])
('EN_A_VS_AN', ['an'])
>>> matches[1].ruleId, matches[1].replacements
('TOT_HE', ['to the'])Print a Match object:
>>> print(matches[1])
Line 1, column 51, Rule ID: TOT_HE[1]
Message: Did you mean 'to the'?
Suggestion: to the
...Automatically apply suggestions to the text:
>>> tool.correct(text)
'A sentence with an error in the Hitchhiker’s Guide to the Galaxy'From the command line:
$ echo 'This are bad.' > example.txt
$ language_tool_python example.txt
example.txt:1:1: THIS_NNS[3]: Did you mean 'these'?language_tool_python runs a LanguageTool Java server in the background. It will shut the server off when garbage collected, for example when a created language_tool_python.LanguageTool object goes out of scope. However, if garbage collection takes awhile, the process might not get deleted right away. If you're seeing lots of processes get spawned and not get deleted, you can explicitly close them:
import language_tool_python
tool = language_tool_python.LanguageToolPublicAPI('de-DE') # starts a process
# do stuff with `tool`
tool.close() # explicitly shut off the LanguageToolYou can also use a context manager (with .. as) to explicitly control when the server is started and stopped:
import language_tool_python
with language_tool_python.LanguageToolPublicAPI('de-DE') as tool:
# do stuff with `tool`
# no need to call `close() as it will happen at the end of the with statementYou can run LanguageTool on one host and connect to it from another. This is useful in some distributed scenarios. Here's a simple example:
>>> import language_tool_python
>>> tool = language_tool_python.LanguageTool('en-US', host='0.0.0.0')
>>> tool._url
'http://0.0.0.0:8081/v2/'>>> import language_tool_python
>>> lang_tool = language_tool_python.LanguageTool('en-US', remote_server='http://0.0.0.0:8081')
>>>
>>>
>>> lang_tool.check('helo darknes my old frend')
[Match({'ruleId': 'UPPERCASE_SENTENCE_START', 'message': 'This sentence does not start with an uppercase letter.', 'replacements': ['Helo'], 'offsetInContext': 0, 'context': 'helo darknes my old frend', 'offset': 0, 'errorLength': 4, 'category': 'CASING', 'ruleIssueType': 'typographical', 'sentence': 'helo darknes my old frend'}), Match({'ruleId': 'MORFOLOGIK_RULE_EN_US', 'message': 'Possible spelling mistake found.', 'replacements': ['darkness', 'darkens', 'darkies'], 'offsetInContext': 5, 'context': 'helo darknes my old frend', 'offset': 5, 'errorLength': 7, 'category': 'TYPOS', 'ruleIssueType': 'misspelling', 'sentence': 'helo darknes my old frend'}), Match({'ruleId': 'MORFOLOGIK_RULE_EN_US', 'message': 'Possible spelling mistake found.', 'replacements': ['friend', 'trend', 'Fred', 'freed', 'Freud', 'Friend', 'fend', 'fiend', 'frond', 'rend', 'fr end'], 'offsetInContext': 20, 'context': 'helo darknes my old frend', 'offset': 20, 'errorLength': 5, 'category': 'TYPOS', 'ruleIssueType': 'misspelling', 'sentence': 'helo darknes my old frend'})]
>>>LanguageTool offers lots of built-in configuration options.
Here's an example of using the configuration options to enable caching. Some users have reported that this helps performance a lot.
import language_tool_python
tool = language_tool_python.LanguageTool('en-US', config={ 'cacheSize': 1000, 'pipelineCaching': True })Here's an example showing how to configure LanguageTool to set a maximum length on grammar-checked text. Will throw an error (which propagates to Python as a language_tool_python.LanguageToolError) if text is too long.
import language_tool_python
tool = language_tool_python.LanguageTool('en-US', config={ 'maxTextLength': 100 })Here's a full list of configuration options. See the LanguageTool HTTPServerConfig documentation for details.
'maxTextLength' - maximum text length, longer texts will cause an error (optional)
'maxTextHardLength' - maximum text length, applies even to users with a special secret 'token' parameter (optional)
'secretTokenKey' - secret JWT token key, if set by user and valid, maxTextLength can be increased by the user (optional)
'maxCheckTimeMillis' - maximum time in milliseconds allowed per check (optional)
'maxErrorsPerWordRate' - checking will stop with error if there are more rules matches per word (optional)
'maxSpellingSuggestions' - only this many spelling errors will have suggestions for performance reasons (optional,
affects Hunspell-based languages only)
'maxCheckThreads' - maximum number of threads working in parallel (optional)
'cacheSize' - size of internal cache in number of sentences (optional, default: 0)
'cacheTTLSeconds' - how many seconds sentences are kept in cache (optional, default: 300 if 'cacheSize' is set)
'requestLimit' - maximum number of requests per requestLimitPeriodInSeconds (optional)
'requestLimitInBytes' - maximum aggregated size of requests per requestLimitPeriodInSeconds (optional)
'timeoutRequestLimit' - maximum number of timeout request (optional)
'requestLimitPeriodInSeconds' - time period to which requestLimit and timeoutRequestLimit applies (optional)
'languageModel' - a directory with '1grams', '2grams', '3grams' sub directories which contain a Lucene index
each with ngram occurrence counts; activates the confusion rule if supported (optional)
'word2vecModel' - a directory with word2vec data (optional), see
https://github.com/languagetool-org/languagetool/blob/master/languagetool-standalone/CHANGES.md#word2vec
'fasttextModel' - a model file for better language detection (optional), see
https://fasttext.cc/docs/en/language-identification.html
'fasttextBinary' - compiled fasttext executable for language detection (optional), see
https://fasttext.cc/docs/en/support.html
'maxWorkQueueSize' - reject request if request queue gets larger than this (optional)
'rulesFile' - a file containing rules configuration, such as .langugagetool.cfg (optional)
'warmUp' - set to 'true' to warm up server at start, i.e. run a short check with all languages (optional)
'blockedReferrers' - a comma-separated list of HTTP referrers (and 'Origin' headers) that are blocked and will not be served (optional)
'premiumOnly' - activate only the premium rules (optional)
'disabledRuleIds' - a comma-separated list of rule ids that are turned off for this server (optional)
'pipelineCaching' - set to 'true' to enable caching of internal pipelines to improve performance
'maxPipelinePoolSize' - cache size if 'pipelineCaching' is set
'pipelineExpireTimeInSeconds' - time after which pipeline cache items expire
'pipelinePrewarming' - set to 'true' to fill pipeline cache on start (can slow down start a lot)
To install via pip:
$ pip install --upgrade language_tool_pythonTo overwrite the host part of URL that is used to download LanguageTool-{version}.zip:
$ export LTP_DOWNLOAD_HOST = [alternate URL]This can be used to downgrade to an older version, for example, or to download from a mirror.
And to choose the specific folder to download the server to:
$ export LTP_PATH = /path/to/save/language/toolThe default download path is ~/.cache/language_tool_python/. The LanguageTool server is about 200 MB, so take that into account when choosing your download folder. (Or, if you you can't spare the disk space, use a remote URL!)
- Python 3.6+
- LanguageTool (Java 8.0 or higher)
The installation process should take care of downloading LanguageTool (it may
take a few minutes). Otherwise, you can manually download
LanguageTool-stable.zip and unzip it
into where the language_tool_python package resides.
As of April 2020, language_tool_python was forked from language-check and no longer supports LanguageTool versions lower than 4.0.
This is a fork of https://github.com/myint/language-check/ that produces more easily parsable results from the command-line.