v0.0.15 🌈

⭐ Major Highlights

• Replaced Gnew package with GoogleNews as it was hampering dependencies upgrade
• Upgrade to Pydantic V2 version
• Removing Python 3.7 support
• Upgraded all dependencies to current latest and greated

Changes

• Use GoogleNews instead of GNews and dependencies upgrade @lalitpagaria (#319)

v0.0.14 🌈

⭐ Major Highlights

• Making dependencies more strict
• Updating more information on Readme
• Adding code coverage reporting with CI job

Changes

• Making dependencies strict @lalitpagaria (#270)

v0.0.13 🌈

⭐ Major Highlights

• Python 3.10 support
• Segregated optional dependencies to support install on need basis to reduce the docker image size
• Added about oraika our parent organization and user

🚀 Features

• Segregated optional dependencies to support install on need basis @GirishPatel (#257)
• Add Python 3.10 support and fix website @lalitpagaria (#218)

🐛 Bug Fixes

• Fixing sample UI dependencies @lalitpagaria (#262)
• Fix UI docker build @lalitpagaria (#255)

🧰 Maintenance

• Updating Readme and fixing pypi release @lalitpagaria (#267)
• Fixing mypy reported issues @lalitpagaria (#268)
• Bump actions/checkout from 2 to 3.1.0 @dependabot (#263)
• Bump actions/cache from 2 to 3.0.11 @dependabot (#265)
• Bump docker/metadata-action from 3.6.2 to 4.1.1 @dependabot (#266)
• Adding numpy to conda build @lalitpagaria (#256)

⚠️Breaking Changes

• To add ElasticSearch 8.x support few input params are modified (#268)
• By default it will only install bare minimal dependencies, in order to install all use pip install obsei[all] (#257)

v0.0.12 🌈

🐛 Bug Fixes

• Fix outscrapper map review API along with moving to faster V3 API @lalitpagaria (#250)
• Fix TwitterSource username bug, add Gnews proxy @chxlium (#246)

🧰 Maintenance

• Upgrade dependencies and add dateparser in dependency list @lalitpagaria (#252)
• Bump actions/setup-python from 2 to 4 @dependabot (#248)
• Bump crazy-max/ghaction-docker-meta from 1 to 3.6.2 @dependabot (#226)
• Upgraded click version to fix typer dependency @GirishPatel (#245)

🚀 Misc

• Add Utilize.app to the list of companies using Obsei @arorajatin (#249)

Thanks to new contributors @chxlium and @arorajatin

v0.0.11 🦚

⭐ Major Highlights

• Youtube: Now fetch Youtube video comments (via Scrapper)
• License: Removed all strong copyleft dependencies
• Demo: Improved demo UI along with adding more detailed logging
• Few bug fixes, dependencies upgrade, CI enhancements and fixing security issue

🚀 Features

• Youtube integration via scrapper @lalitpagaria (#224)
• Removing third party dependencies with strong copyleft licenses @lalitpagaria (#221)
• Enhancing demo UI @lalitpagaria (#214)

🐛 Bug Fixes

• Fixing typing-extensions dependency issue on python 3.7 @lalitpagaria (#217)
• Google News max result fix @tanish36 (#211)
• Bug: Updating long_term_token param to access_token for facebook source. @lalitpagaria (#210)

🧰 Maintenance

• Updated the README @kuutsav (#222)
• Fix security issue with lxml @lalitpagaria (#219)
• Dep upgrade (to address Dependabot for NLTK as well) @lalitpagaria (#215)
• Enabling CI caching @lalitpagaria (#213)

v0.0.10 🌈

⭐ Major Highlights

• Google Maps: Now observer able to fetch google maps reviews
• Handle Long Text: Now use TextSplitter and InferenceAggregator to seamlessly process very long text for example news article
• Analyzer: New TextClassification analyzer let you use non zero-shot classification model
• Analyzer: New Spacy powered NER analyzer let you use spacy based NER models
• Pandas: Now you can use Pandas DataFrame as an observer and informer, which enable you to load and store data from CSV, TSV, Excel and SQL DBs.
• Miscellaneous: Jinja template support for Slack message, new tutorials and pre-commit hook to save dev time

🚀 Features

• Jinja Template for Slack, RegEx & Lammatizer cleaner functions and Sentence based text splitting @lalitpagaria (#206)
• Add SDK and UI-Demo image release job @lalitpagaria (#205)
• Adding jinja template support for slack messages @lalitpagaria (#199)
• Adding google maps review observer via outscrapper api @lalitpagaria (#195)
• add text classification analyzer @shahrukhx01 (#191)
• Add Pandas as Observer/Source @cnarte (#184)
• Adding app_url support to appstore and playstore scrappers @lalitpagaria (#180)
• Adding InferenceAggregator @akar5h (#166)
• spacy ner analyzer , #165 enhancement @akar5h (#171)
• Adding tutorials information and articles @lalitpagaria (#168)
• Colab Tutorials Added @reenabapna (#167)
• TextSplitter Preprocessor Pipeline @akar5h (#160)
• Pre commit integration @salilmishra23 (#156)
• add requirement files for development @salilmishra23 (#159)
• Facebook source time range improvement @GirishPatel (#157)

🐛 Bug Fixes

• Reverting to older messaging format for Twitter v2 API @lalitpagaria (#198)
• Fixing import issues along @lalitpagaria (#193)
• Fix Twitter Source Config import issue @lalitpagaria (#190)
• Use of BaseSettings Causing regression so reverting the changes @lalitpagaria (#189)
• Creating quoted query before passing to GNews client @lalitpagaria (#181)
• Handle Null case when crawler failed to fetch article @lalitpagaria (#173)
• Email source fixed, no duplicates in each iteration @namanjuneja771 (#158)

🧰 Maintenance

• Adding support to dailyget message api @lalitpagaria (#179)

⚠️Breaking Changes

• Fix Twitter Source Config import issue @lalitpagaria (#190)
• Adding InferenceAggregator @akar5h (#166)
• spacy ner analyzer , #165 enhancement @akar5h (#171)

🙏 Release Contributors! ❤️

@akar5h @cnarte @GirishPatel @lalitpagaria @namanjuneja771 @reenabapna @salilmishra23 @shahrukhx01

🥳 New Contributors

• @namanjuneja771 made their first contribution in #158
• @salilmishra23 made their first contribution in #159
• @reenabapna made their first contribution in #167
• @cnarte made their first contribution in #184

Full Changelog: 0.0.9...v0.0.10

Release 0.0.9

⭐ Major Highlights

Pre-Processing (#110)

This added new component to Obsei. It's main idea to provide simple but configurable step to pre-process text before sending it for model prediction. Currently TextCleaning step is added, which helps user to clean raw text's. It is great contribution by @shahrukhx01

Facebook Integration (#150)

Obsei can now observe comments from Facebook page's posts thanks to contribution by @GirishPatel

Google News Integration (#111)

Obsei can now search news on GoogleNews and scrap full news article in text

Website Scrapper Integration (#111)

Obsei can now scrap particular URL or full website if it contains sitemap

Pandas DataFrame Integration (#107)

Obsei can now have Pandas DataFrame as Informer to publish Analyzer's data to DataFrame.

🔆 Other Changes

• Added contribution guideline and code of conduct (e102f89 and fcc9a91)
• Adding version tag along with default logging config (#149)
• Ignoring error during cleaning and fixing exception in google news module (#143)
• Add analyzer batching (#118)
• Adding app name support for app and play store (#128)
• Mypy integration (#135 and #125
• Fixing error regarding offset-naive and offset-aware datetimes comparison (#119)
• Remove import from configuration.py as it is causing loop of import (#133)
• Trimming excessive text before passing to model (#114)
• Ignoring error during cleaning and fixing exception in google news module (#143)
• Adding version tag along with default logging config (#149)
• [BUG] Tokenizer loading error the NER Analyzer (#152)

⚠️ Breaking Changes

• Unifying analyzer request and response (#148)

🙏 Release Contributors! ❤️

@GirishPatel @shahrukhx01 @akar5h @lalitpagaria

Release 0.0.8

This release includes (Refer for detailed changes https://github.com/lalitpagaria/obsei/projects/6) -

1. New Analyzers: Personal Information Anonymizer and Translation
2. GPU support: Analyzer's now can utilize GPU
3. Conda release: Adding initial support to install package from Conda as well as creation of development environment
4. Window support: Adding support for windows platform
5. Obsei UI demo: Adding streamlit based UI to try Obsei
6. Various bug fixes

Release 0.0.7

This release include -

• Email Observer: Currently do not segment attachment (linked issue #30)
• Zendesk Informer (linked issue #31)
• Add extensive example, colab and binder support in Readme (linked issue #39)
• Remove hydra dependency (linked issue #33)
• Adding Logger Informer, so user can easily test out end-to-end pipeline (linked commit ca99da8)
• Adding security policy to repo (linked commit 2685fff)
• [Bug/Regression]: Rest interface via docker build was failing (linked commit ef48b36)

Release 0.0.6

This release include following bug fixes -

• Play store scrapper is failing (linked issue #34)
• Correct analyzer usages in examples (linked issue #36)