xinference 1.11.0__py3-none-any.whl → 1.12.0__py3-none-any.whl

xinference/ui/gradio/utils/latex.py

@@ -0,0 +1,342 @@
+# Copyright 2022-2025 XProbe Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+LaTeX processing utilities for OCR text formatting.
+
+This module provides functions to process LaTeX formulas in OCR text,
+making them compatible with different output formats like Markdown,
+HTML, and pure LaTeX.
+"""
+
+import re
+from typing import Any, Dict, Union
+
+
+def process_latex_formulas(text: str, output_format: str = "markdown") -> str:
+    """
+    Process LaTeX formulas in OCR text to make them compatible with different output formats.
+
+    Args:
+        text: The OCR text containing LaTeX formulas
+        output_format: Target format ("markdown", "html", "latex", "gradio")
+
+    Returns:
+        Processed text with formulas converted to the target format
+    """
+    if not text:
+        return text
+
+    processed_text = text
+
+    if output_format == "markdown":
+        # Convert \[ ... \] to $$ ... $$ for block math in Markdown
+        processed_text = re.sub(
+            r"\\\[\s*(.*?)\s*\\\]",
+            lambda m: f"\n$$\n{m.group(1).strip()}\n$$\n",
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+        # Convert \( ... \) to $ ... $ for inline math in Markdown
+        processed_text = re.sub(
+            r"\\\((.*?)\\\)",
+            lambda m: f"${m.group(1).strip()}$",
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+        # Handle common LaTeX math environments
+        # Convert \begin{equation} ... \end{equation} to $$ ... $$
+        processed_text = re.sub(
+            r"\\begin\{equation\}(.*?)\\end\{equation\}",
+            lambda m: f"\n$$\n{m.group(1).strip()}\n$$\n",
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+        # Convert \begin{align} ... \end{align} to $$ ... $$
+        processed_text = re.sub(
+            r"\\begin\{align\*?\}(.*?)\\end\{align\*?\}",
+            lambda m: f"\n$$\n{m.group(1).strip()}\n$$\n",
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+        # Convert \begin{gather} ... \end{gather} to $$ ... $$
+        processed_text = re.sub(
+            r"\\begin\{gather\}(.*?)\\end\{gather\}",
+            lambda m: f"\n$$\n{m.group(1).strip()}\n$$\n",
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+    elif output_format == "html":
+        # Convert \[ ... \] to <div>...</div> with proper formatting
+        processed_text = re.sub(
+            r"\\\[(.*?)\\\]",
+            lambda m: f'<div class="math-display">\\[{m.group(1).strip()}\\]</div>',
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+        # Convert \( ... \) to <span>...</span> for inline math
+        processed_text = re.sub(
+            r"\\\((.*?)\\\)",
+            lambda m: f'<span class="math-inline">\\({m.group(1).strip()}\\)</span>',
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+    elif output_format == "latex":
+        # Keep original LaTeX format, just clean up spacing
+        processed_text = re.sub(
+            r"\\\[(.*?)\\\]",
+            lambda m: f"\\[\n{m.group(1).strip()}\n\\]",
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+        processed_text = re.sub(
+            r"\\\((.*?)\\\)",
+            lambda m: f"\\({m.group(1).strip()}\\)",
+            processed_text,
+            flags=re.DOTALL,
+        )
+
+    return processed_text
+
+
+def clean_latex_syntax(text: str) -> str:
+    """
+    Clean up common LaTeX syntax issues in OCR text.
+
+    Args:
+        text: The OCR text containing LaTeX
+
+    Returns:
+        Cleaned LaTeX text
+    """
+    if not text:
+        return text
+
+    # Fix common OCR errors in LaTeX
+    cleaned = text
+
+    # Fix spacing around operators
+    cleaned = re.sub(r"(\w)([+\-=*/])(\w)", r"\1 \2 \3", cleaned)
+
+    # Fix double backslashes that might be mangled, but preserve LaTeX delimiters
+    # Only fix excessive backslashes (3+), not double backslashes which are valid LaTeX
+    cleaned = re.sub(r"\\\\{3,}", r"\\\\", cleaned)
+
+    # Fix fractions that might be incorrectly spaced
+    cleaned = re.sub(
+        r"\\frac\s*\{\s*(\w+)\s*\}\s*\{\s*(\w+)\s*\}", r"\\frac{\1}{\2}", cleaned
+    )
+
+    # Fix superscripts and subscripts
+    cleaned = re.sub(r"\^\s*\{\s*(\w+)\s*\}", r"^{\1}", cleaned)
+    cleaned = re.sub(r"_\s*\{\s*(\w+)\s*\}", r"_{\1}", cleaned)
+
+    # Fix common misrecognized Greek letters
+    greek_corrections = {
+        r"\\alpha\s": r"\\alpha ",
+        r"\\beta\s": r"\\beta ",
+        r"\\gamma\s": r"\\gamma ",
+        r"\\delta\s": r"\\delta ",
+        r"\\epsilon\s": r"\\epsilon ",
+        r"\\theta\s": r"\\theta ",
+        r"\\lambda\s": r"\\lambda ",
+        r"\\mu\s": r"\\mu ",
+        r"\\pi\s": r"\\pi ",
+        r"\\sigma\s": r"\\sigma ",
+        r"\\phi\s": r"\\phi ",
+        r"\\omega\s": r"\\omega ",
+        r"\\Delta\s": r"\\Delta ",
+        r"\\Sigma\s": r"\\Sigma ",
+        r"\\Pi\s": r"\\Pi ",
+        r"\\infty\s": r"\\infty ",
+        r"\\pm\s": r"\\pm ",
+        r"\\times\s": r"\\times ",
+        r"\\div\s": r"\\div ",
+        r"\\neq\s": r"\\neq ",
+        r"\\leq\s": r"\\leq ",
+        r"\\geq\s": r"\\geq ",
+        r"\\approx\s": r"\\approx ",
+        r"\\in\s": r"\\in ",
+        r"\\subset\s": r"\\subset ",
+        r"\\supset\s": r"\\supset ",
+        r"\\int\s": r"\\int ",
+        r"\\sum\s": r"\\sum ",
+        r"\\prod\s": r"\\prod ",
+        r"\\partial\s": r"\\partial ",
+        r"\\nabla\s": r"\\nabla ",
+        r"\\sqrt\s": r"\\sqrt ",
+        r"^\{2\}": r"²",
+        r"^\{3\}": r"³",
+    }
+
+    for latex, unicode in greek_corrections.items():
+        cleaned = re.sub(latex, unicode, cleaned)
+
+    return cleaned
+
+
+def process_ocr_latex(text: str, output_format: str = "markdown") -> str:
+    """
+    Convenience function to process OCR text containing LaTeX formulas.
+
+    This function combines LaTeX syntax cleaning and formula formatting
+    in a single call, which is useful for OCR post-processing.
+
+    Args:
+        text: The OCR text containing LaTeX formulas
+        output_format: Target format ("markdown", "html", "latex", "gradio")
+
+    Returns:
+        Processed text with cleaned LaTeX syntax and properly formatted formulas
+    """
+    if not text:
+        return text
+
+    # First clean up LaTeX syntax issues
+    cleaned_text = clean_latex_syntax(text)
+
+    # Then convert formulas to the desired format
+    processed_text = process_latex_formulas(cleaned_text, output_format)
+
+    return processed_text
+
+
+def contains_latex(text: str) -> bool:
+    """
+    Detect if text contains LaTeX formulas.
+
+    Args:
+        text: The text to check for LaTeX formulas
+
+    Returns:
+        True if the text contains LaTeX formulas, False otherwise
+    """
+    if not text:
+        return False
+
+    # Check for common LaTeX indicators
+    latex_indicators = [
+        r"\\\[.*?\\\]",  # Block formulas \[...\]
+        r"\\\(.*?\\\)",  # Inline formulas \(...\)
+        r"\$.*?\$",  # Inline formulas $...$
+        r"\$\$.*?\$\$",  # Block formulas $$...$$
+        r"\\begin\{.*?\}",  # LaTeX environments
+        r"\\[a-zA-Z]+\{",  # LaTeX commands with arguments
+        r"\\[a-zA-Z]+",  # Simple LaTeX commands
+    ]
+
+    for pattern in latex_indicators:
+        if re.search(pattern, text):
+            return True
+
+    return False
+
+
+def process_ocr_result_with_latex(
+    result: Union[str, Dict[str, Any]],
+    output_format: str = "markdown",
+    debug_info: bool = False,
+) -> Union[str, Dict[str, Any]]:
+    """
+    Process OCR results, applying LaTeX formatting if needed.
+
+    This is a unified function to handle both LaTeX detection and processing
+    for OCR results, supporting both string and dict formats.
+
+    Args:
+        result: OCR result, either as string or dict with 'text' key
+        output_format: Target format ("markdown", "html", "latex", "gradio")
+        debug_info: Whether to print debug information about LaTeX processing
+
+    Returns:
+        Processed result with LaTeX formulas formatted appropriately
+    """
+    # Handle dict format (from visualize_ocr and other methods)
+    if isinstance(result, dict):
+        if "text" not in result:
+            return result
+
+        text = result["text"]
+        if not contains_latex(text):
+            return result
+
+        # Process LaTeX
+        processed_text = process_ocr_latex(text, output_format)
+
+        # Create new result dict with processed text
+        processed_result = result.copy()
+        processed_result["text"] = processed_text
+        processed_result["latex_processing"] = {
+            "latex_detected": True,
+            "output_format": output_format,
+        }
+
+        if debug_info:
+            print("🧮 LaTeX Formula Processing:")
+            original_block_formulas = len(re.findall(r"\\\[(.*?)\\\]", text, re.DOTALL))
+            original_inline_dollar = len(re.findall(r"\$(.*?)\$", text, re.DOTALL))
+            original_inline_paren = len(re.findall(r"\\\((.*?)\\\)", text, re.DOTALL))
+            converted_inline = len(re.findall(r"\$(.*?)\$", processed_text, re.DOTALL))
+
+            print(f"  - Original block formulas (\\[...\\]): {original_block_formulas}")
+            print(f"  - Original inline formulas ($...$): {original_inline_dollar}")
+            print(f"  - Original inline formulas (\\(...\\)): {original_inline_paren}")
+            print(f"  - Final inline formulas ($...$): {converted_inline}")
+            print(
+                f"  - Format: Markdown-compatible ($...$ for inline, $$...$$ for block)"
+            )
+
+        return processed_result
+
+    # Handle string format
+    else:
+        # Ensure result is treated as string
+        text_result = str(result)
+
+        if not contains_latex(text_result):
+            return result
+
+        if debug_info:
+            print("🧮 LaTeX Formula Processing:")
+            original_block_formulas = len(
+                re.findall(r"\\\[(.*?)\\\]", text_result, re.DOTALL)
+            )
+            original_inline_dollar = len(
+                re.findall(r"\$(.*?)\$", text_result, re.DOTALL)
+            )
+            original_inline_paren = len(
+                re.findall(r"\\\((.*?)\\\)", text_result, re.DOTALL)
+            )
+
+        processed_text = process_ocr_latex(text_result, output_format)
+
+        if debug_info:
+            converted_inline = len(re.findall(r"\$(.*?)\$", processed_text, re.DOTALL))
+            print(f"  - Original block formulas (\\[...\\]): {original_block_formulas}")
+            print(f"  - Original inline formulas ($...$): {original_inline_dollar}")
+            print(f"  - Original inline formulas (\\(...\\)): {original_inline_paren}")
+            print(f"  - Final inline formulas ($...$): {converted_inline}")
+            print(
+                f"  - Format: Markdown-compatible ($...$ for inline, $$...$$ for block)"
+            )
+
+        return processed_text

xinference/ui/web/ui/build/asset-manifest.json

@@ -1,14 +1,14 @@
 {
   "files": {
     "main.css": "./static/css/main.5ea97072.css",
-    "main.js": "./static/js/main.45e78536.js",
+    "main.js": "./static/js/main.87d6859b.js",
     "static/media/icon.webp": "./static/media/icon.4603d52c63041e5dfbfd.webp",
     "index.html": "./index.html",
     "main.5ea97072.css.map": "./static/css/main.5ea97072.css.map",
-    "main.45e78536.js.map": "./static/js/main.45e78536.js.map"
+    "main.87d6859b.js.map": "./static/js/main.87d6859b.js.map"
   },
   "entrypoints": [
     "static/css/main.5ea97072.css",
-    "static/js/main.45e78536.js"
+    "static/js/main.87d6859b.js"
   ]
 }

xinference/ui/web/ui/build/index.html

@@ -1 +1 @@
-<!doctype html><html lang="en"><head><meta charset="utf-8"/><link rel="icon" href="./favicon.svg"/><meta name="viewport" content="width=device-width,initial-scale=1"/><meta name="theme-color" content="#000000"/><meta name="description" content="Web site created using create-react-app"/><link rel="apple-touch-icon" href="./logo192.png"/><link rel="manifest" href="./manifest.json"/><title>Xinference</title><script defer="defer" src="./static/js/main.45e78536.js"></script><link href="./static/css/main.5ea97072.css" rel="stylesheet"></head><body><noscript>You need to enable JavaScript to run this app.</noscript><div id="root"></div></body></html>
+<!doctype html><html lang="en"><head><meta charset="utf-8"/><link rel="icon" href="./favicon.svg"/><meta name="viewport" content="width=device-width,initial-scale=1"/><meta name="theme-color" content="#000000"/><meta name="description" content="Web site created using create-react-app"/><link rel="apple-touch-icon" href="./logo192.png"/><link rel="manifest" href="./manifest.json"/><title>Xinference</title><script defer="defer" src="./static/js/main.87d6859b.js"></script><link href="./static/css/main.5ea97072.css" rel="stylesheet"></head><body><noscript>You need to enable JavaScript to run this app.</noscript><div id="root"></div></body></html>