html-to-markdown 2.3.0__tar.gz → 2.3.3__tar.gz

{html_to_markdown-2.3.0 → html_to_markdown-2.3.3}/Cargo.lock

@@ -200,9 +200,9 @@ dependencies = [
 
 [[package]]
 name = "clap"
-version = "4.5.48"
+version = "4.5.49"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e2134bb3ea021b78629caa971416385309e0131b351b25e01dc16fb54e1b5fae"
+checksum = "f4512b90fa68d3a9932cea5184017c5d200f5921df706d45e853537dea51508f"
 dependencies = [
  "clap_builder",
  "clap_derive",
@@ -210,9 +210,9 @@ dependencies = [
 
 [[package]]
 name = "clap_builder"
-version = "4.5.48"
+version = "4.5.49"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c2ba64afa3c0a6df7fa517765e31314e983f51dda798ffba27b988194fb65dc9"
+checksum = "0025e98baa12e766c67ba13ff4695a887a1eba19569aad00a472546795bd6730"
 dependencies = [
  "anstream",
  "anstyle",
@@ -222,18 +222,18 @@ dependencies = [
 
 [[package]]
 name = "clap_complete"
-version = "4.5.58"
+version = "4.5.59"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "75bf0b32ad2e152de789bb635ea4d3078f6b838ad7974143e99b99f45a04af4a"
+checksum = "2348487adcd4631696ced64ccdb40d38ac4d31cae7f2eec8817fcea1b9d1c43c"
 dependencies = [
  "clap",
 ]
 
 [[package]]
 name = "clap_derive"
-version = "4.5.47"
+version = "4.5.49"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bbfd7eae0b0f1a6e63d4b13c9c478de77c2eb546fba158ad50b4203dc24b9f9c"
+checksum = "2a0b5487afeab2deb2ff4e03a807ad1a03ac532ff5a2cee5d86884440c7f7671"
 dependencies = [
  "heck",
  "proc-macro2",
@@ -243,15 +243,15 @@ dependencies = [
 
 [[package]]
 name = "clap_lex"
-version = "0.7.5"
+version = "0.7.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b94f61472cee1439c0b966b47e3aca9ae07e45d070759512cd390ea2bebc6675"
+checksum = "a1d728cc89cf3aee9ff92b05e62b19ee65a02b5702cff7d5a377e32c6ae29d8d"
 
 [[package]]
 name = "clap_mangen"
-version = "0.2.29"
+version = "0.2.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "27b4c3c54b30f0d9adcb47f25f61fcce35c4dd8916638c6b82fbd5f4fb4179e2"
+checksum = "263c8214a8e0cb8129f3c62036c50e9c6e15c7bd364c42e0437c492b9293f778"
 dependencies = [
  "clap",
  "roff",
@@ -583,9 +583,9 @@ dependencies = [
 
 [[package]]
 name = "half"
-version = "2.7.0"
+version = "2.7.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e54c115d4f30f52c67202f079c5f9d8b49db4691f460fdb0b4c2e838261b2ba5"
+checksum = "6ea2d84b969582b4b1864a92dc5d27cd2b77b622a8d79306834f1be5ba20d84b"
 dependencies = [
  "cfg-if",
  "crunchy",
@@ -615,7 +615,7 @@ dependencies = [
 
 [[package]]
 name = "html-to-markdown-cli"
-version = "2.3.0"
+version = "2.3.3"
 dependencies = [
  "assert_cmd",
  "clap",
@@ -629,7 +629,7 @@ dependencies = [
 
 [[package]]
 name = "html-to-markdown-node"
-version = "2.3.0"
+version = "2.3.3"
 dependencies = [
  "html-to-markdown-rs",
  "mimalloc-rust",
@@ -640,7 +640,7 @@ dependencies = [
 
 [[package]]
 name = "html-to-markdown-py"
-version = "2.3.0"
+version = "2.3.3"
 dependencies = [
  "base64",
  "html-to-markdown-rs",
@@ -650,7 +650,7 @@ dependencies = [
 
 [[package]]
 name = "html-to-markdown-rs"
-version = "2.3.0"
+version = "2.3.3"
 dependencies = [
  "ammonia",
  "base64",
@@ -667,7 +667,7 @@ dependencies = [
 
 [[package]]
 name = "html-to-markdown-wasm"
-version = "2.3.0"
+version = "2.3.3"
 dependencies = [
  "console_error_panic_hook",
  "getrandom 0.2.16",
@@ -1435,9 +1435,9 @@ dependencies = [
 
 [[package]]
 name = "regex"
-version = "1.12.1"
+version = "1.12.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4a52d8d02cacdb176ef4678de6c052efb4b3da14b78e4db683a4252762be5433"
+checksum = "843bc0191f75f3e22651ae5f1e72939ab2f72a4bc30fa80a066bd66edefc24d4"
 dependencies = [
  "aho-corasick",
  "memchr",
@@ -1447,9 +1447,9 @@ dependencies = [
 
 [[package]]
 name = "regex-automata"
-version = "0.4.12"
+version = "0.4.13"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "722166aa0d7438abbaa4d5cc2c649dac844e8c56d82fb3d33e9c34b5cd268fc6"
+checksum = "5276caf25ac86c8d810222b3dbb938e512c55c6831a10f3e6ed1c93b84041f1c"
 dependencies = [
  "aho-corasick",
  "memchr",
@@ -1458,9 +1458,9 @@ dependencies = [
 
 [[package]]
 name = "regex-syntax"
-version = "0.8.7"
+version = "0.8.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c3160422bbd54dd5ecfdca71e5fd59b7b8fe2b1697ab2baf64f6d05dcc66d298"
+checksum = "7a2d987857b319362043e95f5353c0535c1f58eec5336fdfcf626430af7def58"
 
 [[package]]
 name = "roff"

{html_to_markdown-2.3.0 → html_to_markdown-2.3.3}/Cargo.toml

@@ -3,7 +3,7 @@ resolver = "2"
 members = ["crates/html-to-markdown-py"]
 
 [workspace.package]
-version = "2.3.0"
+version = "2.3.3"
 edition = "2021"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
 license = "MIT"
@@ -15,7 +15,7 @@ rust-version = "1.80"
 
 [workspace.dependencies]
 # Core library
-html-to-markdown-rs = { version = "2.3.0", path = "crates/html-to-markdown" }
+html-to-markdown-rs = { version = "2.3.3", path = "crates/html-to-markdown" }
 
 # HTML parsing and sanitization
 tl = "0.7"

{html_to_markdown-2.3.0 → html_to_markdown-2.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: html-to-markdown
-Version: 2.3.0
+Version: 2.3.3
 Classifier: Development Status :: 5 - Production/Stable
 Classifier: Environment :: Console
 Classifier: Intended Audience :: Developers

{html_to_markdown-2.3.0 → html_to_markdown-2.3.3}/crates/html-to-markdown/src/hocr/converter.rs

@@ -156,8 +156,15 @@ fn convert_element(
 
         // Paragraphs
         HocrElementType::OcrPar => {
-            if !output.is_empty() && !output.ends_with("\n\n") {
-                output.push_str("\n\n");
+            let bullet_paragraph = is_bullet_paragraph(element);
+            if !output.is_empty() {
+                if bullet_paragraph {
+                    if !output.ends_with('\n') {
+                        output.push('\n');
+                    }
+                } else if !output.ends_with("\n\n") {
+                    output.push_str("\n\n");
+                }
             }
 
             if let Some(heading) = detect_heading_paragraph(element) {
@@ -188,7 +195,13 @@ fn convert_element(
             if output.ends_with(' ') {
                 output.pop();
             }
-            output.push_str("\n\n");
+            if bullet_paragraph {
+                if !output.ends_with('\n') {
+                    output.push('\n');
+                }
+            } else {
+                output.push_str("\n\n");
+            }
         }
 
         // Blockquotes
@@ -588,6 +601,43 @@ fn try_spatial_table_reconstruction(element: &HocrElement) -> Option<String> {
     None
 }
 
+fn is_bullet_paragraph(element: &HocrElement) -> bool {
+    if element.element_type != HocrElementType::OcrPar {
+        return false;
+    }
+
+    let text = element_text_content(element);
+    let trimmed = text.trim_start();
+    if trimmed.is_empty() {
+        return false;
+    }
+
+    if matches!(trimmed.chars().next(), Some('•' | '●' | '-' | '+' | '*')) {
+        return true;
+    }
+
+    let mut chars = trimmed.chars().peekable();
+    let mut digit_count = 0;
+    while let Some(&ch) = chars.peek() {
+        if ch.is_ascii_digit() {
+            digit_count += 1;
+            chars.next();
+        } else {
+            break;
+        }
+    }
+
+    if digit_count > 0 {
+        if let Some(&ch) = chars.peek() {
+            if (ch == '.' || ch == ')') && chars.clone().nth(1).map(|c| c.is_whitespace()).unwrap_or(false) {
+                return true;
+            }
+        }
+    }
+
+    false
+}
+
 #[derive(Clone)]
 struct CodeLineInfo {
     text: String,
@@ -662,6 +712,10 @@ fn collect_code_block(children: &[&HocrElement]) -> Option<(Vec<String>, usize,
         return None;
     }
 
+    if !is_confident_code_block(&collected) {
+        return None;
+    }
+
     // Determine base indentation metrics
     let mut x_values: Vec<u32> = collected
         .iter()
@@ -844,6 +898,42 @@ fn is_bullet_like(line: &str) -> bool {
     false
 }
 
+fn contains_keyword_token(text: &str, keyword: &str) -> bool {
+    text.split(|ch: char| !(ch.is_ascii_alphanumeric() || ch == '_'))
+        .any(|token| token == keyword)
+}
+
+fn is_shell_prompt(text: &str) -> bool {
+    let trimmed = text.trim_start();
+    if trimmed.is_empty() {
+        return false;
+    }
+
+    trimmed.starts_with('$')
+        || trimmed.starts_with('#')
+        || trimmed.contains("]#")
+        || trimmed.starts_with("sudo ")
+        || trimmed.starts_with("./")
+        || trimmed.starts_with("python ")
+        || trimmed.starts_with("pip ")
+        || trimmed.starts_with("uv ")
+}
+
+fn starts_with_keyword(trimmed: &str, keyword: &str) -> bool {
+    if !trimmed.starts_with(keyword) {
+        return false;
+    }
+    if let Some(first) = trimmed.chars().next() {
+        if !first.is_ascii_lowercase() {
+            return false;
+        }
+    }
+    match trimmed.chars().nth(keyword.len()) {
+        None => true,
+        Some(ch) => ch.is_whitespace() || matches!(ch, '(' | ':' | '{' | '[' | '.'),
+    }
+}
+
 fn is_code_paragraph(lines: &[CodeLineInfo]) -> bool {
     if lines.is_empty() {
         return false;
@@ -865,47 +955,70 @@ fn is_code_paragraph(lines: &[CodeLineInfo]) -> bool {
 
         total += 1;
         let lower = text.to_lowercase();
+        let trimmed = text.trim_start();
 
-        let has_keyword = lower.contains("function")
-            || lower.contains("console.")
-            || lower.starts_with("return")
-            || lower.starts_with("var ")
-            || lower.starts_with("let ")
-            || lower.starts_with("const ")
-            || lower.starts_with("async ")
-            || lower.starts_with("await ")
-            || lower.starts_with("if ")
-            || lower.starts_with("elif ")
-            || lower.starts_with("else if ")
-            || lower.starts_with("for ")
-            || lower.starts_with("while ")
-            || lower.starts_with("switch ")
-            || lower.starts_with("case ")
-            || lower.starts_with("class ")
-            || lower.starts_with("struct ")
-            || lower.starts_with("enum ")
-            || lower.starts_with("def ")
-            || lower.starts_with("fn ")
-            || lower.starts_with("pub ")
-            || lower.starts_with("import ")
-            || lower.starts_with("from ")
-            || lower.starts_with("using ")
-            || lower.starts_with("namespace ");
-
-        let has_symbol = text.contains('{')
-            || text.contains('}')
-            || text.contains(';')
-            || text.contains("::")
-            || text.contains("->")
-            || text.contains("=>");
+        let documentation_tokens = [
+            "definition",
+            "theorem",
+            "lemma",
+            "proof",
+            "corollary",
+            "algorithm",
+            "figure",
+            "table",
+            "appendix",
+        ];
+        if documentation_tokens
+            .iter()
+            .any(|token| contains_keyword_token(&lower, token))
+        {
+            return false;
+        }
+
+        let has_keyword = (starts_with_keyword(trimmed, "function") && text.contains('('))
+            || (starts_with_keyword(trimmed, "return")
+                && trimmed
+                    .chars()
+                    .nth("return".len())
+                    .map(|c| c.is_whitespace())
+                    .unwrap_or(true))
+            || trimmed.starts_with("console.")
+            || starts_with_keyword(trimmed, "async")
+            || starts_with_keyword(trimmed, "await")
+            || (starts_with_keyword(trimmed, "class") && (text.contains('{') || text.contains(':')))
+            || (starts_with_keyword(trimmed, "struct") && text.contains('{'))
+            || (starts_with_keyword(trimmed, "enum") && text.contains('{'))
+            || (starts_with_keyword(trimmed, "def") && (text.contains('(') || text.contains(':')))
+            || (starts_with_keyword(trimmed, "fn") && text.contains('('))
+            || (starts_with_keyword(trimmed, "pub")
+                && (text.contains("fn") || text.contains("struct") || text.contains("enum")))
+            || starts_with_keyword(trimmed, "import")
+            || starts_with_keyword(trimmed, "using")
+            || starts_with_keyword(trimmed, "namespace")
+            || starts_with_keyword(trimmed, "public")
+            || starts_with_keyword(trimmed, "private")
+            || starts_with_keyword(trimmed, "protected")
+            || starts_with_keyword(trimmed, "static")
+            || starts_with_keyword(trimmed, "void")
+            || starts_with_keyword(trimmed, "try")
+            || starts_with_keyword(trimmed, "catch")
+            || starts_with_keyword(trimmed, "finally")
+            || starts_with_keyword(trimmed, "throw")
+            || starts_with_keyword(trimmed, "typedef")
+            || starts_with_keyword(trimmed, "package")
+            || starts_with_keyword(trimmed, "module");
+
+        let has_symbol = text.contains(';') || text.contains("::");
 
         if has_keyword || has_symbol {
             strong_markers += 1;
             continue;
         }
 
-        let trimmed = text.trim_start();
-        let starts_with_indent = trimmed.len() + 2 <= text.len();
+        if is_shell_prompt(text) {
+            strong_markers += 1;
+            continue;
+        }
         let has_assignment = text.contains(" = ")
             || text.contains("+=")
             || text.contains("-=")
@@ -914,7 +1027,11 @@ fn is_code_paragraph(lines: &[CodeLineInfo]) -> bool {
             || text.contains(" := ")
             || text.contains(" == ");
 
-        if has_assignment || starts_with_indent {
+        let has_arrow = text.contains("=>");
+        let has_brace = text.contains('{') || text.contains('}');
+        let has_pointer_arrow = text.contains("->");
+
+        if has_assignment || has_arrow || has_brace || has_pointer_arrow {
             moderate_markers += 1;
         }
     }
@@ -922,10 +1039,13 @@ fn is_code_paragraph(lines: &[CodeLineInfo]) -> bool {
     if total == 0 {
         return false;
     }
+    if strong_markers == 0 {
+        return false;
+    }
     if strong_markers * 2 >= total {
         return true;
     }
-    strong_markers > 0 && (strong_markers + moderate_markers) * 2 >= total
+    (strong_markers + moderate_markers) * 2 >= total
 }
 
 fn normalize_code_line(text: &str) -> String {
@@ -980,6 +1100,114 @@ fn normalize_code_line(text: &str) -> String {
     final_line.trim().to_string()
 }
 
+fn is_confident_code_block(lines: &[CodeLineInfo]) -> bool {
+    let mut total = 0;
+    let mut keyword_lines = 0;
+    let mut punctuation_lines = 0;
+    let mut assignment_lines = 0;
+    let mut shell_lines = 0;
+    let mut indent_lines = 0;
+
+    let min_x = lines.iter().map(|info| info.x1).min().unwrap_or_default();
+
+    for info in lines {
+        let text = info.text.trim();
+        if text.is_empty() {
+            continue;
+        }
+        total += 1;
+
+        if is_shell_prompt(text) {
+            shell_lines += 1;
+        }
+
+        let trimmed = text.trim_start();
+
+        if (starts_with_keyword(trimmed, "function") && text.contains('('))
+            || trimmed.starts_with("console.")
+            || (starts_with_keyword(trimmed, "return")
+                && trimmed
+                    .chars()
+                    .nth("return".len())
+                    .map(|c| c.is_whitespace())
+                    .unwrap_or(true))
+            || starts_with_keyword(trimmed, "async")
+            || starts_with_keyword(trimmed, "await")
+            || (starts_with_keyword(trimmed, "class") && (text.contains('{') || text.contains(':')))
+            || (starts_with_keyword(trimmed, "struct") && text.contains('{'))
+            || (starts_with_keyword(trimmed, "enum") && text.contains('{'))
+            || (starts_with_keyword(trimmed, "def") && (text.contains('(') || text.contains(':')))
+            || (starts_with_keyword(trimmed, "fn") && text.contains('('))
+            || (starts_with_keyword(trimmed, "pub")
+                && (text.contains("fn") || text.contains("struct") || text.contains("enum")))
+            || starts_with_keyword(trimmed, "import")
+            || starts_with_keyword(trimmed, "using")
+            || starts_with_keyword(trimmed, "namespace")
+            || starts_with_keyword(trimmed, "public")
+            || starts_with_keyword(trimmed, "private")
+            || starts_with_keyword(trimmed, "protected")
+            || starts_with_keyword(trimmed, "static")
+            || starts_with_keyword(trimmed, "void")
+            || starts_with_keyword(trimmed, "try")
+            || starts_with_keyword(trimmed, "catch")
+            || starts_with_keyword(trimmed, "finally")
+            || starts_with_keyword(trimmed, "throw")
+            || starts_with_keyword(trimmed, "typedef")
+            || starts_with_keyword(trimmed, "package")
+            || starts_with_keyword(trimmed, "module")
+        {
+            keyword_lines += 1;
+        }
+
+        if text.contains(';')
+            || text.contains('{')
+            || text.contains('}')
+            || text.contains("::")
+            || text.contains("->")
+            || text.contains("=>")
+        {
+            punctuation_lines += 1;
+        }
+
+        if text.contains(" = ")
+            || text.contains("+=")
+            || text.contains("-=")
+            || text.contains("*=")
+            || text.contains("/=")
+            || text.contains(" := ")
+            || text.contains(" == ")
+        {
+            assignment_lines += 1;
+        }
+
+        if info.x1 > min_x + 8 {
+            indent_lines += 1;
+        }
+    }
+
+    if total < 3 {
+        return false;
+    }
+
+    if shell_lines >= 2 && shell_lines * 2 >= total {
+        return true;
+    }
+
+    if keyword_lines >= 2 && assignment_lines >= 1 {
+        return true;
+    }
+
+    if keyword_lines >= 1 && punctuation_lines >= 1 && assignment_lines >= 1 {
+        return true;
+    }
+
+    if indent_lines == total && keyword_lines >= 1 && assignment_lines >= 1 {
+        return true;
+    }
+
+    false
+}
+
 fn detect_code_language(lines: &[String]) -> Option<&'static str> {
     let lower_lines: Vec<String> = lines.iter().map(|line| line.to_lowercase()).collect();
     if lower_lines.iter().any(|line| line.contains("function"))

{html_to_markdown-2.3.0 → html_to_markdown-2.3.3}/crates/html-to-markdown-py/README.md

@@ -3,7 +3,7 @@
 High-performance HTML → Markdown conversion powered by Rust. Shipping as a Rust crate, Python package, Node.js bindings, WebAssembly, and standalone CLI with identical rendering behaviour.
 
 [![PyPI version](https://badge.fury.io/py/html-to-markdown.svg)](https://pypi.org/project/html-to-markdown/)
-[![npm version](https://badge.fury.io/js/html-to-markdown.svg)](https://www.npmjs.com/package/html-to-markdown)
+[![npm version](https://badge.fury.io/js/html-to-markdown.svg)](https://www.npmjs.com/package/html-to-markdown-node)
 [![Crates.io](https://img.shields.io/crates/v/html-to-markdown-rs.svg)](https://crates.io/crates/html-to-markdown-rs)
 [![Python Versions](https://img.shields.io/pypi/pyversions/html-to-markdown.svg)](https://pypi.org/project/html-to-markdown/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Goldziher/html-to-markdown/blob/main/LICENSE)
@@ -23,9 +23,9 @@ High-performance HTML → Markdown conversion powered by Rust. Shipping as a Rus
 
 | Target                      | Command                                                                   |
 | --------------------------- | ------------------------------------------------------------------------- |
-| **Node.js/Bun** (native)    | `npm install @html-to-markdown/node`                                      |
-| **WebAssembly** (universal) | `npm install @html-to-markdown/wasm`                                      |
-| **Deno**                    | `import { convert } from "npm:@html-to-markdown/wasm"`                    |
+| **Node.js/Bun** (native)    | `npm install html-to-markdown-node`                                       |
+| **WebAssembly** (universal) | `npm install html-to-markdown-wasm`                                       |
+| **Deno**                    | `import { convert } from "npm:html-to-markdown-wasm"`                     |
 | **Python** (bindings + CLI) | `pip install html-to-markdown`                                            |
 | **Rust** crate              | `cargo add html-to-markdown-rs`                                           |
 | Rust CLI                    | `cargo install html-to-markdown-cli`                                      |
@@ -39,7 +39,7 @@ High-performance HTML → Markdown conversion powered by Rust. Shipping as a Rus
 **Node.js / Bun (Native - Fastest):**
 
 ```typescript
-import { convert } from '@html-to-markdown/node';
+import { convert } from 'html-to-markdown-node';
 
 const html = '<h1>Hello</h1><p>Rust ❤️ Markdown</p>';
 const markdown = convert(html, {
@@ -52,8 +52,8 @@ const markdown = convert(html, {
 **Deno / Browsers / Edge (Universal):**
 
 ```typescript
-import { convert } from "npm:@html-to-markdown/wasm"; // Deno
-// or: import { convert } from '@html-to-markdown/wasm'; // Bundlers
+import { convert } from "npm:html-to-markdown-wasm"; // Deno
+// or: import { convert } from 'html-to-markdown-wasm'; // Bundlers
 
 const markdown = convert(html, {
   headingStyle: 'atx',

{html_to_markdown-2.3.0 → html_to_markdown-2.3.3}/html_to_markdown/__init__.py

@@ -15,7 +15,13 @@ V1 API (backward compatibility):
     markdown = convert_to_markdown(html, heading_style="atx")
 """
 
-from html_to_markdown.api import convert
+from html_to_markdown.api import (
+    InlineImage,
+    InlineImageConfig,
+    InlineImageWarning,
+    convert,
+    convert_with_inline_images,
+)
 from html_to_markdown.exceptions import (
     ConflictingOptionsError,
     EmptyHtmlError,
@@ -31,12 +37,16 @@ __all__ = [
     "ConversionOptions",
     "EmptyHtmlError",
     "HtmlToMarkdownError",
+    "InlineImage",
+    "InlineImageConfig",
+    "InlineImageWarning",
     "InvalidParserError",
     "MissingDependencyError",
     "PreprocessingOptions",
     "convert",
     "convert_to_markdown",
+    "convert_with_inline_images",
     "markdownify",
 ]
 
-__version__ = "2.3.0"
+__version__ = "2.3.3"

html_to_markdown-2.3.3/html_to_markdown/api.py

@@ -0,0 +1,143 @@
+"""New v2 functional API for HTML to Markdown conversion.
+
+This module provides the new functional API with dataclass-based options,
+using the Rust backend for conversion.
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Literal, TypedDict, cast
+
+import html_to_markdown._html_to_markdown as _rust  # type: ignore[import-not-found]
+from html_to_markdown.options import ConversionOptions, PreprocessingOptions
+
+if TYPE_CHECKING:
+    from html_to_markdown._html_to_markdown import InlineImageConfig
+else:
+    InlineImageConfig = _rust.InlineImageConfig  # type: ignore[misc, assignment]
+
+
+class InlineImage(TypedDict):
+    """Inline image extracted during conversion."""
+
+    data: bytes
+    format: str
+    filename: str | None
+    description: str | None
+    dimensions: tuple[int, int] | None
+    source: Literal["img_data_uri", "svg_element"]
+    attributes: dict[str, str]
+
+
+class InlineImageWarning(TypedDict):
+    """Warning produced during inline image extraction."""
+
+    index: int
+    message: str
+
+
+def _to_rust_preprocessing(options: PreprocessingOptions) -> _rust.PreprocessingOptions:
+    """Convert high-level preprocessing options to the Rust bindings."""
+    return _rust.PreprocessingOptions(
+        enabled=options.enabled,
+        preset=options.preset,
+        remove_navigation=options.remove_navigation,
+        remove_forms=options.remove_forms,
+    )
+
+
+def _to_rust_options(
+    options: ConversionOptions,
+    preprocessing: PreprocessingOptions,
+) -> _rust.ConversionOptions:
+    """Convert high-level conversion options to the Rust bindings."""
+    return _rust.ConversionOptions(
+        heading_style=options.heading_style,
+        list_indent_type=options.list_indent_type,
+        list_indent_width=options.list_indent_width,
+        bullets=options.bullets,
+        strong_em_symbol=options.strong_em_symbol,
+        escape_asterisks=options.escape_asterisks,
+        escape_underscores=options.escape_underscores,
+        escape_misc=options.escape_misc,
+        escape_ascii=options.escape_ascii,
+        code_language=options.code_language,
+        autolinks=options.autolinks,
+        default_title=options.default_title,
+        br_in_tables=options.br_in_tables,
+        hocr_spatial_tables=options.hocr_spatial_tables,
+        highlight_style=options.highlight_style,
+        extract_metadata=options.extract_metadata,
+        whitespace_mode=options.whitespace_mode,
+        strip_newlines=options.strip_newlines,
+        wrap=options.wrap,
+        wrap_width=options.wrap_width,
+        convert_as_inline=options.convert_as_inline,
+        sub_symbol=options.sub_symbol,
+        sup_symbol=options.sup_symbol,
+        newline_style=options.newline_style,
+        code_block_style=options.code_block_style,
+        keep_inline_images_in=list(options.keep_inline_images_in) if options.keep_inline_images_in else [],
+        preprocessing=_to_rust_preprocessing(preprocessing),
+        encoding=options.encoding,
+        debug=options.debug,
+        strip_tags=list(options.strip_tags) if options.strip_tags else [],
+    )
+
+
+def convert(
+    html: str,
+    options: ConversionOptions | None = None,
+    preprocessing: PreprocessingOptions | None = None,
+) -> str:
+    """Convert HTML to Markdown using the Rust backend.
+
+    Args:
+        html: HTML string to convert.
+        options: Conversion configuration options (defaults to ConversionOptions()).
+        preprocessing: HTML preprocessing options (defaults to PreprocessingOptions()).
+
+    Returns:
+        Converted Markdown string.
+    """
+    if options is None:
+        options = ConversionOptions()
+    if preprocessing is None:
+        preprocessing = PreprocessingOptions()
+
+    rust_options = _to_rust_options(options, preprocessing)
+    return cast("str", _rust.convert(html, rust_options))
+
+
+def convert_with_inline_images(
+    html: str,
+    options: ConversionOptions | None = None,
+    preprocessing: PreprocessingOptions | None = None,
+    image_config: InlineImageConfig | None = None,
+) -> tuple[str, list[InlineImage], list[InlineImageWarning]]:
+    """Convert HTML and extract inline images.
+
+    Returns Markdown along with extracted inline images and any warnings.
+    """
+    if options is None:
+        options = ConversionOptions()
+    if preprocessing is None:
+        preprocessing = PreprocessingOptions()
+    if image_config is None:
+        image_config = InlineImageConfig()
+
+    rust_options = _to_rust_options(options, preprocessing)
+    markdown, images, warnings = cast(
+        "tuple[str, list[InlineImage], list[InlineImageWarning]]",
+        _rust.convert_with_inline_images(html, rust_options, image_config),
+    )
+    return markdown, list(images), list(warnings)
+
+
+__all__ = [
+    "InlineImage",
+    "InlineImageConfig",
+    "InlineImageWarning",
+    "convert",
+    "convert_with_inline_images",
+]

{html_to_markdown-2.3.0 → html_to_markdown-2.3.3}/pyproject.toml

@@ -7,7 +7,7 @@ requires = [
 
 [project]
 name = "html-to-markdown"
-version = "2.3.0"
+version = "2.3.3"
 description = "High-performance HTML to Markdown converter powered by Rust with a clean Python API"
 readme = "README_PYPI.md"
 keywords = [

html_to_markdown-2.3.0/html_to_markdown/api.py

@@ -1,74 +0,0 @@
-"""New v2 functional API for HTML to Markdown conversion.
-
-This module provides the new functional API with dataclass-based options,
-using the Rust backend for conversion.
-"""
-
-from __future__ import annotations
-
-import html_to_markdown._html_to_markdown as _rust  # type: ignore[import-not-found]
-from html_to_markdown.options import ConversionOptions, PreprocessingOptions
-
-
-def convert(
-    html: str,
-    options: ConversionOptions | None = None,
-    preprocessing: PreprocessingOptions | None = None,
-) -> str:
-    """Convert HTML to Markdown using the Rust backend.
-
-    Args:
-        html: HTML string to convert.
-        options: Conversion configuration options (defaults to ConversionOptions()).
-        preprocessing: HTML preprocessing options (defaults to PreprocessingOptions()).
-
-    Returns:
-        Converted Markdown string.
-    """
-    if options is None:
-        options = ConversionOptions()
-    if preprocessing is None:
-        preprocessing = PreprocessingOptions()
-
-    rust_preprocessing = _rust.PreprocessingOptions(
-        enabled=preprocessing.enabled,
-        preset=preprocessing.preset,
-        remove_navigation=preprocessing.remove_navigation,
-        remove_forms=preprocessing.remove_forms,
-    )
-
-    rust_options = _rust.ConversionOptions(
-        heading_style=options.heading_style,
-        list_indent_type=options.list_indent_type,
-        list_indent_width=options.list_indent_width,
-        bullets=options.bullets,
-        strong_em_symbol=options.strong_em_symbol,
-        escape_asterisks=options.escape_asterisks,
-        escape_underscores=options.escape_underscores,
-        escape_misc=options.escape_misc,
-        escape_ascii=options.escape_ascii,
-        code_language=options.code_language,
-        autolinks=options.autolinks,
-        default_title=options.default_title,
-        br_in_tables=options.br_in_tables,
-        hocr_spatial_tables=options.hocr_spatial_tables,
-        highlight_style=options.highlight_style,
-        extract_metadata=options.extract_metadata,
-        whitespace_mode=options.whitespace_mode,
-        strip_newlines=options.strip_newlines,
-        wrap=options.wrap,
-        wrap_width=options.wrap_width,
-        convert_as_inline=options.convert_as_inline,
-        sub_symbol=options.sub_symbol,
-        sup_symbol=options.sup_symbol,
-        newline_style=options.newline_style,
-        code_block_style=options.code_block_style,
-        keep_inline_images_in=list(options.keep_inline_images_in) if options.keep_inline_images_in else [],
-        preprocessing=rust_preprocessing,
-        encoding=options.encoding,
-        debug=options.debug,
-        strip_tags=list(options.strip_tags) if options.strip_tags else [],
-    )
-
-    result: str = _rust.convert(html, rust_options)
-    return result