add default english selection and new languages #27
Conversation
WalkthroughThe pull request expands the report language selector in Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~8 minutes
Poem
Pre-merge checks and finishing touches✅ Passed checks (3 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
Tip 📝 Customizable high-level summaries are now available in beta!You can now customize how CodeRabbit generates the high-level summary in your pull requests — including its content, structure, tone, and formatting.
Example instruction:
Note: This feature is currently in beta for Pro-tier users, and pricing will be announced later. Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
levelup/app.py(1 hunks)
🔇 Additional comments (1)
levelup/app.py (1)
331-335: LGTM! Default language selection implemented correctly.The use of
index=language_options.index("English")correctly sets English as the default selection. The implementation is clean and straightforward.
| language_options = [ | ||
| "Czech", | ||
| "Danish", | ||
| "Dutch", | ||
| "English", | ||
| "German", | ||
| "Finnish", | ||
| "French", | ||
| "German", | ||
| "Greek", | ||
| "Italian", | ||
| "Kurdish (Kurmanji)", | ||
| "Polish", | ||
| "Portuguese", | ||
| "Russian", | ||
| "Turkish", | ||
| "Spanish", | ||
| "Swedish", | ||
| "Turkish", | ||
| "Ukrainian", | ||
| ] |
There was a problem hiding this comment.
🧩 Analysis chain
Verify LLM support for all language options, especially dialect specifications.
The alphabetical ordering and expanded language list improve UX. However, ensure that the Gemini 2.0 Flash Lite model supports all these languages, particularly "Kurdish (Kurmanji)" with the parenthetical dialect specification, as LLMs may expect standardized language codes or names.
🏁 Script executed:
#!/bin/bash
# Verify if the prompt construction uses these language names directly
# and check for any language normalization logic
rg -n -A5 -B5 'def get_resume_analysis_prompt' --type=py
# Check if there are any language mappings or validations
rg -n 'language.*map|lang.*code|ISO.*639' --type=pyLength of output: 441
🌐 Web query:
Does Google Gemini 2.0 Flash Lite support Kurdish Kurmanji language and what is the correct way to specify it?
💡 Result:
Short answer: No — Gemini 2.0 / Gemini Flash Lite does not list Kurdish (Kurmanji) among its supported languages, so specifying a Kurmanji tag will not be recognized by the model. [1]
If you still need to try a language tag, use a BCP‑47-style tag for Kurmanji such as "kmr" or a Kurdish tag with script subtag (e.g. "ku-Latn" for Kurdish in Latin script). Those are the correct forms if/when the service accepts Kurdish. [2][3]
Sources:
- Gemini Flash / Flash‑Lite TTS docs (available languages list). [1]
- Examples and guidance on BCP‑47 language tags for Kurdish / Kurmanji ("kmr", "ku-Latn", etc.). [2][3]
🏁 Script executed:
#!/bin/bash
# Inspect how report_language is used in the prompt
cat levelup/prompts.py | head -50
# Search for API calls that use the language parameter
rg -n -A3 -B3 'report_language|language.*format|Gemini|genai\.generate' --type=py -C3Length of output: 5257
🏁 Script executed:
#!/bin/bash
# Check the full app.py to see how language is passed to the prompt function
rg -n -A10 'get_resume_analysis_prompt' levelup/app.pyLength of output: 889
🌐 Web query:
Google Gemini 2.0 Flash Lite supported languages list documentation
💡 Result:
Here’s the official supported-languages information for Gemini 2.0 Flash‑Lite (from Google docs):
Supported languages (can understand and respond in):
Arabic (ar), Bengali (bn), Bulgarian (bg), Chinese — simplified & traditional (zh), Croatian (hr), Czech (cs), Danish (da), Dutch (nl), English (en), Estonian (et), Finnish (fi), French (fr), German (de), Greek (el), Hebrew (iw), Hindi (hi), Hungarian (hu), Indonesian (id), Italian (it), Japanese (ja), Korean (ko), Latvian (lv), Lithuanian (lt), Norwegian (no), Polish (pl), Portuguese (pt), Romanian (ro), Russian (ru), Serbian (sr), Slovak (sk), Slovenian (sl), Spanish (es), Swahili (sw), Swedish (sv), Thai (th), Turkish (tr), Ukrainian (uk), Vietnamese (vi). [1][2]
Additional languages supported by Gemini 2.0 Flash / Flash-Lite family (extended list includes): Afrikaans, Amharic, Assamese, Azerbaijani, Belarusian, Bosnian, Catalan, Cebuano, Corsican, Welsh, Dhivehi, Esperanto, Basque, Persian, Filipino (Tagalog), Frisian, Irish, Scots Gaelic, Galician, Gujarati, Hausa, Hawaiian, Hmong, Haitian Creole, Armenian, Igbo, Icelandic, Javanese, Georgian, Kazakh, Khmer, Kannada, Krio, Kurdish, Kyrgyz, Latin, Luxembourgish, Lao, Malagasy, Maori, Macedonian, Malayalam, Mongolian, Meiteilon (Manipuri), Marathi, Malay, Maltese, Burmese, Nepali, Nyanja (Chichewa), Odia (Oriya), Punjabi, Pashto, Sindhi, Sinhala, Samoan, Shona, Somali, Albanian, Sesotho, Sundanese, Tamil, Telugu, Tajik, Uyghur, Urdu, Uzbek, Xhosa, Yiddish, Yoruba, Zulu. [1][3]
Sources / docs:
- Vertex AI / Gemini 2.0 Flash‑Lite docs (supported languages referenced on model page). [1]
- Google AI / Gemini models supported-languages page. [2]
- Firebase / Vertex-AI supported-models languages summary. [3]
Would you like direct links to those documentation pages or a machine-readable (CSV/JSON) version of the lists?
🌐 Web query:
Gemini API language support ISO 639 BCP 47 language codes
💡 Result:
Short answer: Google’s Gemini APIs accept standard BCP‑47 language tags (which are built from ISO‑639 language codes plus optional script/region). Gemini model docs list supported ISO‑639 two‑letter codes (e.g. en, fr, zh, hi, etc.), and Gemini‑TTS publishes specific BCP‑47 locale codes (e.g. en-US, fr-FR, ja-JP). See Google Cloud docs for the full supported-language lists. [1][2]
References
- Gemini models — supported languages (ISO‑639 / BCP‑47 tags listing). [1]
- Gemini‑TTS — available languages with BCP‑47 locale codes (e.g. en‑US, nl‑NL, ja‑JP). [2]
[1] https://cloud.google.com/vertex-ai/generative-ai/docs/models.md
[2] https://cloud.google.com/text-to-speech/docs/gemini-tts
🌐 Web query:
Does Gemini support language names in English or require language codes
💡 Result:
Use language codes. Gemini's API docs list supported languages with ISO/BCP-47 codes (e.g., "en", "zh", "en-US"/"fr-FR" for TTS); API parameters expect those codes (UIs may show names). [1][2]
Sources:
[1] Gemini models — Google AI developer docs.
[2] Gemini TTS / Cloud Text-to-Speech — Google Cloud docs.
Remove or correct "Kurdish (Kurmanji)" — unsupported format for Gemini 2.0 Flash Lite.
Gemini 2.0 Flash Lite does not recognize "Kurdish (Kurmanji)" with the parenthetical dialect specification. The supported languages list includes "Kurdish" but not the "(Kurmanji)" variant. All other 16 languages in the list (Czech through Ukrainian) are supported.
Action: Either remove "Kurdish (Kurmanji)" from the language options, or replace it with "Kurdish" (without parentheses). Additionally, consider mapping language names to ISO-639 codes (e.g., "da" for Danish, "nl" for Dutch) rather than English names to follow API conventions.
🤖 Prompt for AI Agents
In levelup/app.py around lines 312 to 330, the language option "Kurdish
(Kurmanji)" is in an unsupported format for Gemini 2.0 Flash Lite; replace that
entry with "Kurdish" (no parentheses) or remove it entirely to match the model's
supported list, and as a follow-up prefer using a mapping of language display
names to ISO-639 codes (e.g., {"Danish": "da", "Dutch": "nl", "Kurdish": "ku",
...}) so the API sends standardized codes instead of plain English names.
Summary by CodeRabbit
Release Notes