Repository navigation
Releases
Bhash Jupyter kernel 0.1.16 — keyword and construct lookup
Version 0.1.16 adds offline keyword and construct lookup, alongside the standard-library lookup to JupyterLab 4.2–4.x and Jupyter Notebook 7.2–7.x. It bundles compiler 0.28.0, the same verified Linux x64, Windows x64 and Apple Silicon compression providers, and the Hindi starter/compression notebooks. A full JDK 17+ is required to execute Bhash cells. End users do not need Maven, Node.js/npm, C development tools or a JupyterLab rebuild.
हिन्दी: नया ⌨ शब्द/ढाँचे toolbar नियंत्रण खोलें या Command Palette में Bhash: Keywords and Constructs · शब्द और ढाँचे चुनें। 57 grammar/contextual शब्द और 25 संपादन योग्य ढाँचे हिन्दी/English खोज, विषय फ़िल्टर और कोड preview के साथ मिलते हैं। शब्द के पास के identifiers अलग रहते हैं; कई पंक्तियों के ढाँचे वर्तमान indentation रखते हैं और पहला placeholder चुनते हैं। Undo से पूरा संपादन वापस लें। कक्ष अपने-आप नहीं चलता।
The Library/Language buttons switch between both tools in one reference pane. Templates cover functions, main, classes, records, interfaces, variables, conditions, loops, resources, exceptions, async/await, tagged unions and pattern matching. Replace placeholders and supply the enclosing scope/imports needed by the construct before running it. Notebook 7.2 sidebar closing/reopening is also fixed.
हिन्दी: नोटबुक की toolbar में 📚 पुस्तकालय दबाएँ, या Command Palette में Bhash: Standard Library Lookup · मानक पुस्तकालय खोलें। 131 प्रकार और 940 सदस्यों में हिन्दी या अंग्रेज़ी नाम से खोजें, हस्ताक्षर और उदाहरण पढ़ें, फिर चुने code कक्ष में कॉल डालें। पहला तर्क चुन जाता है; अपने मान भरें। Undo से संपादन वापस लें। उदाहरण नया code कक्ष बनाता है। कक्ष अपने-आप नहीं चलता।
The lookup supports Hindi/English search, classes and functions, overloads, parameters, return types, target restrictions, examples, light/dark themes and narrow panes. Changed, closed, read-only and non-Bhash cells are protected. The catalog and examples are included for offline use; full documentation links open online. Hindi lookup tutorial.
Install or upgrade / स्थापना और अद्यतन
Run these commands in the Python environment that starts the Jupyter server. Before registration, record any custom settings you want to keep.
If you registered a custom compiler, Java executable or database drivers, repeat your previous --compiler, --java, --mysql-driver and --sqlserver-driver options when running python -m bhash_jupyter.install. --compiler overrides the bundled compiler; repeat it only if you want to keep using your own JAR. Registration replaces the kernelspec, so omitting those options may remove the earlier settings. Save downloaded notebooks under new names to preserve existing work.
हिन्दी: पंजीकरण से पहले अपना पुराना --compiler, --java, --mysql-driver और --sqlserver-driver विकल्प लिख लें और नीचे पंजीकरण करते समय फिर दें। --compiler अपना JAR चुनता है; शामिल संकलक चाहिए तो वह विकल्प न दें।
python -m pip install --upgrade "bhash-jupyter[lab] @ https://github.com/bhashitva/bhash/releases/download/jupyter-0.1.16/bhash_jupyter-0.1.16-py3-none-any.whl"
python -m bhash_jupyter.install
python -m pip show bhash-jupyterFor Notebook 7, replace [lab] with [notebook] in the first command. Download hindi-starter.ipynb and compression.ipynb from this release; both are also included in the wheel. SHA256SUMS covers the wheel, notebooks and release-info.json. The provenance file records the exact source revision and bundled compiler/catalog hashes; the release tag points to that source revision.
हिन्दी: पैकेज केवल किसी दूसरे कर्नेल के Python वातावरण में लगाने से browser के नियंत्रण नहीं बदलेंगे। Jupyter सर्वर वाले वातावरण में लगाएँ। Jupyter सर्वर बंद करके दोबारा चलाएँ और browser पृष्ठ reload करें; केवल कर्नेल restart करना पर्याप्त नहीं है। फिर Bhash (भाष) कर्नेल चुनें। pip show में 0.1.16 दिखना चाहिए। python -m jupyter labextension list में @bhashitva/library-lookup v0.1.16 enabled दिखेगा।
हिन्दी: पुराना Notebook 6/NbClassic यह browser विस्तार नहीं चलाता। VS Code की notebook UI भी JupyterLab विस्तार नहीं लोड करती; वहाँ Bhash VS Code extension का Bhash: Standard Library Lookup उपयोग करें। JupyterLab में lookup अलग पैनल में और Notebook 7 में दाएँ sidebar में खुलता है।
Try it / पहला प्रयोग
Search class, वर्ग, foreach or प्रत्येक in ⌨ शब्द/ढाँचे. Use Keywords/Constructs and topic filters, select a result, inspect its preview and insert it into the selected Bhash code cell. print under Constructs inserts लिखो(); with the cursor inside the parentheses; fill it with "नमस्ते!" and run the cell yourself.
हिन्दी: सूची और उदाहरण offline हैं। छोटे पैनल में परिणाम चुनने पर विवरण खुलता है; ← खोज के नतीजे से वापस जाएँ। केवल शब्द, पूरा ढाँचा या पुस्तकालय की कॉल डालना source संपादन है; उससे कोई कक्ष अपने-आप नहीं चलता।
Search absolute or निरपेक्ष, choose the integer overload and insert the call. Replace the selected placeholder with -३, then run the cell yourself:
मानक.गणित.निरपेक्ष(-३)
The result is ३. Search text split, समुच्चय or जेसन to explore other APIs. Example fragments may need the declarations, imports or resource setup described in their reference pages.
Full installed-kernel client smoke tests, all 57 keywords, the Hindi starter and forced-GC compression tests run on Linux, Windows and macOS before publication. The exact shipped wheel also passes browser tests on current and minimum-supported JupyterLab/Notebook versions, including all 57 keywords and 25 constructs, selection replacement, Undo, indentation, example contents, read-only/stale-cell protection, mode switching, reopening, real execution, offline search and responsive themes.
हिन्दी: कर्नेल का भाषा व्यवहार नहीं बदला है। अदिश बंधन, प्रकार और फलन अगले कक्षों में रहते हैं; वस्तु bindings उसी कक्ष तक सीमित हैं। Strict रिक्त-सुरक्षा वैकल्पिक रहती है। पूरी जानकारी कर्नेल मार्गदर्शिका में पढ़ें।
VS Code extension 0.29.0 — standard library lookup
VS Code extension 0.29.0
Open Bhash: Standard Library Lookup from the Command Palette, the editor's book icon, or Bhash Library in the status bar.
Search 131 standard-library types and 940 declared/inherited members in Hindi or English. Browse constructors, overloads, parameters, return types, Hindi descriptions and examples; insert qualified calls/types with editable Tab placeholders and isolated Undo.
The reference shares its catalog and renderer with the Playground, follows VS Code's theme, fits narrow panes, and works offline and in Restricted Mode. Browsing does not need a compiler connection. Source text is never sent to the reference webview. Full tutorial links open in the browser.
This extension retains the verified Bhash compiler 0.28.0, including its bundled compression providers. Compiler, SDK and Jupyter releases are unchanged.
हिन्दी
कमांड पैलेट में Bhash: Standard Library Lookup चलाएँ, संपादक के ऊपर पुस्तक चिह्न दबाएँ या नीचे Bhash Library चुनें।
split, गणित या पूरा मानक.पाठ्य.बाँटो नाम खोजें। निर्माता, ओवरलोड, मापदंड, फल का प्रकार, हिन्दी विवरण और उदाहरण पढ़ें। कॉल / नाम डालें से कोड में पूरा नाम डालें; चुने तर्क भरें और Tab से आगे जाएँ। Undo से परिवर्तन वापस लें।
पुस्तकालय संदर्भ एक्सटेंशन के साथ आता है और offline तथा Restricted Mode में भी उपलब्ध है। संकलन और विश्लेषण के लिए Java की आवश्यकता बनी रहती है। पूरे मार्गदर्शक के लिंक के लिए इंटरनेट चाहिए। यदि lookup खोलने के बाद स्रोत बदल गया हो तो संपादक में नया स्थान चुनकर फिर डालें।
एक्सटेंशन का संस्करण 0.29.0 है; bundled संकलक 0.28.0 बना रहता है।
Bhash VS Code extension 0.28.0
Bundles compiler 0.28.0 with typed C-library interoperability, bounded compression, verified providers for three desktop hosts, GC-safe resource ownership and Hindi documentation. See the changelog.
Bhash Jupyter kernel 0.1.15 — standard-library lookup
Version 0.1.15 adds an offline standard-library lookup to JupyterLab 4.2–4.x and Jupyter Notebook 7.2–7.x. It bundles compiler 0.28.0, the same verified Linux x64, Windows x64 and Apple Silicon compression providers, and the Hindi starter/compression notebooks. A full JDK 17+ is required to execute Bhash cells. End users do not need Maven, Node.js/npm, C development tools or a JupyterLab rebuild.
हिन्दी: नोटबुक की toolbar में 📚 पुस्तकालय दबाएँ, या Command Palette में Bhash: Standard Library Lookup · मानक पुस्तकालय खोलें। 131 प्रकार और 940 सदस्यों में हिन्दी या अंग्रेज़ी नाम से खोजें, हस्ताक्षर और उदाहरण पढ़ें, फिर चुने code कक्ष में कॉल डालें। पहला तर्क चुन जाता है; अपने मान भरें। Undo से संपादन वापस लें। उदाहरण नया code कक्ष बनाता है। कक्ष अपने-आप नहीं चलता।
The lookup supports Hindi/English search, classes and functions, overloads, parameters, return types, target restrictions, examples, light/dark themes and narrow panes. Changed, closed, read-only and non-Bhash cells are protected. The catalog and examples are included for offline use; full documentation links open online. Hindi lookup tutorial.
Install or upgrade / स्थापना और अद्यतन
Run these commands in the Python environment that starts the Jupyter server. Before registration, record any custom settings you want to keep.
If you registered a custom compiler, Java executable or database drivers, repeat your previous --compiler, --java, --mysql-driver and --sqlserver-driver options when running python -m bhash_jupyter.install. --compiler overrides the bundled compiler; repeat it only if you want to keep using your own JAR. Registration replaces the kernelspec, so omitting those options may remove the earlier settings. Save downloaded notebooks under new names to preserve existing work.
हिन्दी: पंजीकरण से पहले अपना पुराना --compiler, --java, --mysql-driver और --sqlserver-driver विकल्प लिख लें और नीचे पंजीकरण करते समय फिर दें। --compiler अपना JAR चुनता है; शामिल संकलक चाहिए तो वह विकल्प न दें।
python -m pip install --upgrade "bhash-jupyter[lab] @ https://github.com/bhashitva/bhash/releases/download/jupyter-0.1.15/bhash_jupyter-0.1.15-py3-none-any.whl"
python -m bhash_jupyter.install
python -m pip show bhash-jupyterFor Notebook 7, replace [lab] with [notebook] in the first command. Download hindi-starter.ipynb and compression.ipynb from this release; both are also included in the wheel. SHA256SUMS covers the wheel, notebooks and release-info.json. The provenance file records the exact source revision and bundled compiler/catalog hashes; the release tag points to that source revision.
हिन्दी: पैकेज केवल किसी दूसरे कर्नेल के Python वातावरण में लगाने से browser के नियंत्रण नहीं बदलेंगे। Jupyter सर्वर वाले वातावरण में लगाएँ। Jupyter सर्वर बंद करके दोबारा चलाएँ और browser पृष्ठ reload करें; केवल कर्नेल restart करना पर्याप्त नहीं है। फिर Bhash (भाष) कर्नेल चुनें। pip show में 0.1.15 दिखना चाहिए। python -m jupyter labextension list में @bhashitva/library-lookup v0.1.15 enabled दिखेगा।
हिन्दी: पुराना Notebook 6/NbClassic यह browser विस्तार नहीं चलाता। VS Code की notebook UI भी JupyterLab विस्तार नहीं लोड करती; वहाँ Bhash VS Code extension का Bhash: Standard Library Lookup उपयोग करें। JupyterLab में lookup अलग पैनल में और Notebook 7 में दाएँ sidebar में खुलता है।
Try it / पहला प्रयोग
Search absolute or निरपेक्ष, choose the integer overload and insert the call. Replace the selected placeholder with -३, then run the cell yourself:
मानक.गणित.निरपेक्ष(-३)
The result is ३. Search text split, समुच्चय or जेसन to explore other APIs. Example fragments may need the declarations, imports or resource setup described in their reference pages.
Full installed-kernel client smoke tests, all 57 keywords, the Hindi starter and forced-GC compression tests run on Linux, Windows and macOS before publication. The exact shipped wheel also passes browser tests on current and minimum-supported JupyterLab/Notebook versions, including selection replacement, Undo, example contents, read-only/stale-cell protection, real execution, offline search and responsive themes.
हिन्दी: कर्नेल का भाषा व्यवहार नहीं बदला है। अदिश बंधन, प्रकार और फलन अगले कक्षों में रहते हैं; वस्तु bindings उसी कक्ष तक सीमित हैं। Strict रिक्त-सुरक्षा वैकल्पिक रहती है। पूरी जानकारी कर्नेल मार्गदर्शिका में पढ़ें।
Bhash Jupyter kernel 0.1.14
Version 0.1.14 bundles compiler 0.28.0 with typed C-library interoperability and bounded zlib/gzip compression. Verified Linux x64 (glibc 2.35+), Windows x64 and macOS 14+ Apple Silicon providers are included. A full JDK 17+ is required; Maven, zlib headers and C development tools are not required for notebook compression.
हिन्दी: मानक.संपीड़न से जेडलिब/गज़िप डेटा बनाएँ और खोलें। सफलता या अपेक्षित विफलता को मानक.परिणाम में सँभालें। प्रदाता कर्नेल में शामिल हैं; code अपने-आप डाउनलोड नहीं होता। हिन्दी संपीड़न अभ्यास पढ़ें।
Install or upgrade in the Python environment where you run Jupyter:
python -m pip install --upgrade "bhash-jupyter[lab] @ https://github.com/bhashitva/bhash/releases/download/jupyter-0.1.14/bhash_jupyter-0.1.14-py3-none-any.whl"
python -m bhash_jupyter.install
python -m pip show bhash-jupyterRestart the notebook kernel and select Bhash (भाष). Download hindi-starter.ipynb and compression.ipynb from this release; the installed wheel includes both. Full installed-kernel smoke tests, including compression, run on Linux, Windows and macOS before publication.
हिन्दी: यदि पहले अलग Java या database drivers चुने थे, python -m bhash_jupyter.install में वही --java, --mysql-driver और --sqlserver-driver विकल्प दोहराएँ। हर पंजीकरण kernelspec बदलता है; विकल्प छोड़ने पर पुरानी सेटिंग हट सकती है। नई नोटबुक को अलग फ़ाइल में सहेजें।
पहले कक्ष में यह उदाहरण चलाएँ:
आयात मानक.संपीड़न;
आयात मानक.संपीड़नप्रारूप;
मिलान (संपीड़न.संपीड़ितकरो(यूटीएफ८बफ़र("नमस्ते भारत 😀"), संपीड़नप्रारूप.गज़िप, ६)) {
विकल्प मानक.परिणाम.सफल(स्थिर पैक): { लिखो("संपीड़न सफल"); }
विकल्प मानक.परिणाम.असफल(स्थिर दोष): { लिखो(दोष.संदेश); }
विकल्प रिक्त: { लिखो("अप्रत्याशित रिक्त परिणाम"); }
}
Scalar bindings, types and functions persist across cells. Object bindings remain cell-local; create and consume buffers/Results in the same cell, or use a persistent function. Strict null safety remains opt-in. Browser/Playground and Android targets do not support C bindings. Native compilation uses the supported SDK; generated native executables remain independent of JVM/.NET.
Typed C providers 1.0.0
Verified zlib 1.4 providers for Linux x64 glibc 2.35+, Windows x64 UCRT and macOS 14+ Apple Silicon. Includes pinned contracts, runtime identity, per-host provenance and redistribution notices. Built only after cross-target and Jupyter validation on every host.
Bhash standalone compiler 0.28.0
Typed C-library interoperability and bounded zlib/gzip compression (मानक.संपीड़न) on JVM, desktop .NET and native. Includes verified Linux x64, Windows x64 and Apple Silicon static/shared providers, typed errors, ownership checks and GC safety. Full JDK 17+ required; native runtime ABI 18 and C-binding ABI 1. Browser/Android C bindings are unsupported. See the Hindi tutorial.
Bhash VS Code extension 0.27.0
Bundles compiler 0.27.0 with tagged unions, exhaustive matching, typed Result, contextual highlighting, case-factory assistance, pattern-binding hover and missing-case Quick Fixes. See the changelog.
Bhash VS Code extension 0.26.0
Bundles compiler 0.26.0 with opt-in null safety, nullable-aware editor assistance, rename/find references and safer everyday programming. See the changelog.
Bhash Jupyter kernel 0.1.13
Version 0.1.13 bundles compiler 0.27.0 with tagged unions, exhaustive pattern matching, typed Result and Unit. Requires a full JDK 17+; Maven is not required. The installed wheel is verified through a real Jupyter client on Linux, Windows and macOS before publication.
हिन्दी: विकल्पसमूह, मिलान, मानक.परिणाम<त, द> और मानक.इकाई अब पहले से बने कर्नेल में मिलते हैं। हिन्दी ट्यूटोरियल पढ़ें।
Install or upgrade in the Python environment where you run Jupyter:
python -m pip install --upgrade "bhash-jupyter[lab] @ https://github.com/bhashitva/bhash/releases/download/jupyter-0.1.13/bhash_jupyter-0.1.13-py3-none-any.whl"
python -m bhash_jupyter.install
python -m pip show bhash-jupyterRestart the notebook kernel and select Bhash (भाष). Re-register external database drivers with --mysql-driver or --sqlserver-driver as before. Download hindi-starter.ipynb directly from the release assets. To copy it with python -m bhash_jupyter.install --notebook Hindi-Starter.ipynb, include the same --java, --mysql-driver and --sqlserver-driver options used for your existing kernel, since every installer invocation replaces the kernelspec. Use a new destination filename; use --java for a JDK outside PATH.
हिन्दी: नोटबुक को installer से कॉपी करते समय पहले चुने गए --java, --mysql-driver और --sqlserver-driver विकल्प उसी आदेश में दोहराएँ। हर बार installer कर्नेल का पंजीकरण बदलता है; विकल्प छोड़ने पर पुराने driver और Java स्थान की सेटिंग हट सकती है। केवल नोटबुक चाहिए तो release assets से सीधे डाउनलोड करें।
पहले कक्ष में यह उदाहरण चलाएँ। Strict profile के लिए directive सबसे पहले रखें:
// bhash:null-safety=strict
स्थिर उत्तर = मानक.परिणाम<पूर्णांक, पाठ>.सफल(७);
मिलान (उत्तर) {
विकल्प मानक.परिणाम.सफल(स्थिर मूल्य) => अंकपाठ(मूल्य),
विकल्प मानक.परिणाम.असफल(स्थिर दोष) => दोष
}
यह ७ देता है। घोषणाएँ और फलन अगले कक्षों में रहते हैं; वस्तु bindings उसी कक्ष तक सीमित हैं। Result को उसी कक्ष में बनाएँ और मिलान करें, या अगले कक्ष में बुलाने के लिए लौटाने वाला फलन घोषित करें। Strict null safety remains opt-in and persists for the notebook session.