rlmgrep 0.1.7__tar.gz → 0.1.9__tar.gz

{rlmgrep-0.1.7 → rlmgrep-0.1.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rlmgrep
-Version: 0.1.7
+Version: 0.1.9
 Summary: Grep-shaped CLI search powered by DSPy RLM
 Author: rlmgrep
 License: MIT
@@ -17,12 +17,13 @@ Grep-shaped search powered by DSPy RLM. It accepts a natural-language query, sca
 ## Quickstart
 
 ```sh
-uv tool install --python 3.11 rlmgrep
+uv tool install rlmgrep
 # or from GitHub:
-# uv tool install --python 3.11 git+https://github.com/halfprice06/rlmgrep.git
+# uv tool install git+https://github.com/halfprice06/rlmgrep.git
 
 export OPENAI_API_KEY=...  # or set keys in ~/.rlmgrep
-rlmgrep "where are API keys read" rlmgrep/
+rlmgrep --answer "What does this repo do and where are the entry points?" .
+rlmgrep -n -C 2 "Where is retry/backoff configured and what are the defaults?" .
 ```
 
 ## Requirements
@@ -31,6 +32,20 @@ rlmgrep "where are API keys read" rlmgrep/
 - Deno runtime (DSPy RLM uses a Deno-based interpreter)
 - API key for your chosen provider (OpenAI, Anthropic, Gemini, etc.)
 
+## Non-text Files (PDF + Office + Media)
+
+One of rlmgrep’s most useful features is that it can “grep” **PDFs and Office files** by converting them into text before the RLM search runs.
+
+How it works:
+- **PDFs** are parsed with `pypdf`. Each page gets a marker line like `===== Page N =====`, and output lines include a `page=N` suffix. Line numbers refer to the extracted text (not PDF coordinates).
+- **Office & binary docs** (`.docx`, `.pptx`, `.xlsx`, `.html`, `.zip`, etc.) are converted to Markdown via **MarkItDown**. This happens during ingestion, so rlmgrep can search them like any other text file.
+- **Images** can be described by a vision model through MarkItDown (OpenAI/Anthropic/Gemini).
+- **Audio** transcription is supported through OpenAI when enabled.
+
+Sidecar caching:
+- For images/audio, converted text is cached next to the original file as `<original>.<ext>.md` and reused on later runs.
+- Use `-a/--text` if you want to treat binary files as raw text (UTF‑8 with replacement) and skip conversion.
+
 ## Install Deno
 
 DSPy requires the Deno runtime. Install it with the official scripts:
@@ -61,6 +76,7 @@ rlmgrep [options] "query" [paths...]
 
 Common options:
 
+- `--answer` return a narrative answer before the grep output
 - `-C N` context lines before/after (grep-style)
 - `-A N` context lines after
 - `-B N` context lines before
@@ -174,13 +190,6 @@ CLI flags override config values. Model keys are resolved as:
 
 If more than one provider key is set and the model does not make the provider obvious, rlmgrep emits a warning and requires an explicit `--api-key`.
 
-## Non-text files (PDF, images, audio)
-
-- PDF files are parsed with `pypdf`. Each page gets a marker line `===== Page N =====`, and output lines include a `page=N` suffix.
-- Images and audio are converted via `markitdown` when enabled in config. Image conversion supports `openai`, `anthropic`, and `gemini` providers; audio conversion currently supports `openai` only.
-- Converted image/audio text is cached in sidecar files named `<original>.<ext>.md` next to the original file and reused on subsequent runs.
-- Use `-a/--text` to force binary files to be read as text (UTF-8 with replacement).
-
 ## Skill (Anthropic-style)
 
 A ready-to-copy skill lives in:

{rlmgrep-0.1.7 → rlmgrep-0.1.9}/README.md

@@ -5,12 +5,13 @@ Grep-shaped search powered by DSPy RLM. It accepts a natural-language query, sca
 ## Quickstart
 
 ```sh
-uv tool install --python 3.11 rlmgrep
+uv tool install rlmgrep
 # or from GitHub:
-# uv tool install --python 3.11 git+https://github.com/halfprice06/rlmgrep.git
+# uv tool install git+https://github.com/halfprice06/rlmgrep.git
 
 export OPENAI_API_KEY=...  # or set keys in ~/.rlmgrep
-rlmgrep "where are API keys read" rlmgrep/
+rlmgrep --answer "What does this repo do and where are the entry points?" .
+rlmgrep -n -C 2 "Where is retry/backoff configured and what are the defaults?" .
 ```
 
 ## Requirements
@@ -19,6 +20,20 @@ rlmgrep "where are API keys read" rlmgrep/
 - Deno runtime (DSPy RLM uses a Deno-based interpreter)
 - API key for your chosen provider (OpenAI, Anthropic, Gemini, etc.)
 
+## Non-text Files (PDF + Office + Media)
+
+One of rlmgrep’s most useful features is that it can “grep” **PDFs and Office files** by converting them into text before the RLM search runs.
+
+How it works:
+- **PDFs** are parsed with `pypdf`. Each page gets a marker line like `===== Page N =====`, and output lines include a `page=N` suffix. Line numbers refer to the extracted text (not PDF coordinates).
+- **Office & binary docs** (`.docx`, `.pptx`, `.xlsx`, `.html`, `.zip`, etc.) are converted to Markdown via **MarkItDown**. This happens during ingestion, so rlmgrep can search them like any other text file.
+- **Images** can be described by a vision model through MarkItDown (OpenAI/Anthropic/Gemini).
+- **Audio** transcription is supported through OpenAI when enabled.
+
+Sidecar caching:
+- For images/audio, converted text is cached next to the original file as `<original>.<ext>.md` and reused on later runs.
+- Use `-a/--text` if you want to treat binary files as raw text (UTF‑8 with replacement) and skip conversion.
+
 ## Install Deno
 
 DSPy requires the Deno runtime. Install it with the official scripts:
@@ -49,6 +64,7 @@ rlmgrep [options] "query" [paths...]
 
 Common options:
 
+- `--answer` return a narrative answer before the grep output
 - `-C N` context lines before/after (grep-style)
 - `-A N` context lines after
 - `-B N` context lines before
@@ -162,13 +178,6 @@ CLI flags override config values. Model keys are resolved as:
 
 If more than one provider key is set and the model does not make the provider obvious, rlmgrep emits a warning and requires an explicit `--api-key`.
 
-## Non-text files (PDF, images, audio)
-
-- PDF files are parsed with `pypdf`. Each page gets a marker line `===== Page N =====`, and output lines include a `page=N` suffix.
-- Images and audio are converted via `markitdown` when enabled in config. Image conversion supports `openai`, `anthropic`, and `gemini` providers; audio conversion currently supports `openai` only.
-- Converted image/audio text is cached in sidecar files named `<original>.<ext>.md` next to the original file and reused on subsequent runs.
-- Use `-a/--text` to force binary files to be read as text (UTF-8 with replacement).
-
 ## Skill (Anthropic-style)
 
 A ready-to-copy skill lives in:

{rlmgrep-0.1.7 → rlmgrep-0.1.9}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "rlmgrep"
-version = "0.1.7"
+version = "0.1.9"
 description = "Grep-shaped CLI search powered by DSPy RLM"
 readme = "README.md"
 requires-python = ">=3.11"

{rlmgrep-0.1.7 → rlmgrep-0.1.9}/rlmgrep/__init__.py

@@ -1,2 +1,2 @@
 __all__ = ["__version__"]
-__version__ = "0.1.7"
+__version__ = "0.1.9"

{rlmgrep-0.1.7 → rlmgrep-0.1.9}/rlmgrep.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rlmgrep
-Version: 0.1.7
+Version: 0.1.9
 Summary: Grep-shaped CLI search powered by DSPy RLM
 Author: rlmgrep
 License: MIT
@@ -17,12 +17,13 @@ Grep-shaped search powered by DSPy RLM. It accepts a natural-language query, sca
 ## Quickstart
 
 ```sh
-uv tool install --python 3.11 rlmgrep
+uv tool install rlmgrep
 # or from GitHub:
-# uv tool install --python 3.11 git+https://github.com/halfprice06/rlmgrep.git
+# uv tool install git+https://github.com/halfprice06/rlmgrep.git
 
 export OPENAI_API_KEY=...  # or set keys in ~/.rlmgrep
-rlmgrep "where are API keys read" rlmgrep/
+rlmgrep --answer "What does this repo do and where are the entry points?" .
+rlmgrep -n -C 2 "Where is retry/backoff configured and what are the defaults?" .
 ```
 
 ## Requirements
@@ -31,6 +32,20 @@ rlmgrep "where are API keys read" rlmgrep/
 - Deno runtime (DSPy RLM uses a Deno-based interpreter)
 - API key for your chosen provider (OpenAI, Anthropic, Gemini, etc.)
 
+## Non-text Files (PDF + Office + Media)
+
+One of rlmgrep’s most useful features is that it can “grep” **PDFs and Office files** by converting them into text before the RLM search runs.
+
+How it works:
+- **PDFs** are parsed with `pypdf`. Each page gets a marker line like `===== Page N =====`, and output lines include a `page=N` suffix. Line numbers refer to the extracted text (not PDF coordinates).
+- **Office & binary docs** (`.docx`, `.pptx`, `.xlsx`, `.html`, `.zip`, etc.) are converted to Markdown via **MarkItDown**. This happens during ingestion, so rlmgrep can search them like any other text file.
+- **Images** can be described by a vision model through MarkItDown (OpenAI/Anthropic/Gemini).
+- **Audio** transcription is supported through OpenAI when enabled.
+
+Sidecar caching:
+- For images/audio, converted text is cached next to the original file as `<original>.<ext>.md` and reused on later runs.
+- Use `-a/--text` if you want to treat binary files as raw text (UTF‑8 with replacement) and skip conversion.
+
 ## Install Deno
 
 DSPy requires the Deno runtime. Install it with the official scripts:
@@ -61,6 +76,7 @@ rlmgrep [options] "query" [paths...]
 
 Common options:
 
+- `--answer` return a narrative answer before the grep output
 - `-C N` context lines before/after (grep-style)
 - `-A N` context lines after
 - `-B N` context lines before
@@ -174,13 +190,6 @@ CLI flags override config values. Model keys are resolved as:
 
 If more than one provider key is set and the model does not make the provider obvious, rlmgrep emits a warning and requires an explicit `--api-key`.
 
-## Non-text files (PDF, images, audio)
-
-- PDF files are parsed with `pypdf`. Each page gets a marker line `===== Page N =====`, and output lines include a `page=N` suffix.
-- Images and audio are converted via `markitdown` when enabled in config. Image conversion supports `openai`, `anthropic`, and `gemini` providers; audio conversion currently supports `openai` only.
-- Converted image/audio text is cached in sidecar files named `<original>.<ext>.md` next to the original file and reused on subsequent runs.
-- Use `-a/--text` to force binary files to be read as text (UTF-8 with replacement).
-
 ## Skill (Anthropic-style)
 
 A ready-to-copy skill lives in: