philoch-bib-sdk 0.3.9__tar.gz → 0.4.0__tar.gz

{philoch_bib_sdk-0.3.9 → philoch_bib_sdk-0.4.0}/Cargo.lock

@@ -27,6 +27,37 @@ version = "1.0.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "9330f8b2ff13f34540b44e946ef35111825727b38d33286ef986142615121801"
 
+[[package]]
+name = "crossbeam-deque"
+version = "0.8.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9dd111b7b7f7d55b72c0a6ae361660ee5853c9af73f70c3c2ef6858b950e2e51"
+dependencies = [
+ "crossbeam-epoch",
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "crossbeam-epoch"
+version = "0.9.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5b82ac4a3c2ca9c3460964f020e1402edd5753411d7737aa39c3714ad1b5420e"
+dependencies = [
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "crossbeam-utils"
+version = "0.8.21"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d0a5c400df2834b80a4c3327b3aad3a4c4cd4de0629063962b03235697506a28"
+
+[[package]]
+name = "either"
+version = "1.15.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "48c757948c5ede0e46177b7add2e67155f70e33c07fea8284df6576da70b3719"
+
 [[package]]
 name = "getrandom"
 version = "0.3.4"
@@ -77,10 +108,12 @@ checksum = "42f5e15c9953c5e4ccceeb2e7382a716482c34515315f7b03532b8b4e8393d2d"
 
 [[package]]
 name = "philoch_bib_sdk"
-version = "0.1.6"
+version = "0.4.0"
 dependencies = [
  "ahash",
  "pyo3",
+ "rayon",
+ "strsim",
 ]
 
 [[package]]
@@ -100,11 +133,10 @@ dependencies = [
 
 [[package]]
 name = "pyo3"
-version = "0.22.6"
+version = "0.25.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f402062616ab18202ae8319da13fa4279883a2b8a9d9f83f20dbade813ce1884"
+checksum = "8970a78afe0628a3e3430376fc5fd76b6b45c4d43360ffd6cdd40bdde72b682a"
 dependencies = [
- "cfg-if",
  "indoc",
  "libc",
  "memoffset",
@@ -118,9 +150,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-build-config"
-version = "0.22.6"
+version = "0.25.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b14b5775b5ff446dd1056212d778012cbe8a0fbffd368029fd9e25b514479c38"
+checksum = "458eb0c55e7ece017adeba38f2248ff3ac615e53660d7c71a238d7d2a01c7598"
 dependencies = [
  "once_cell",
  "target-lexicon",
@@ -128,9 +160,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-ffi"
-version = "0.22.6"
+version = "0.25.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9ab5bcf04a2cdcbb50c7d6105de943f543f9ed92af55818fd17b660390fc8636"
+checksum = "7114fe5457c61b276ab77c5055f206295b812608083644a5c5b2640c3102565c"
 dependencies = [
  "libc",
  "pyo3-build-config",
@@ -138,9 +170,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros"
-version = "0.22.6"
+version = "0.25.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0fd24d897903a9e6d80b968368a34e1525aeb719d568dba8b3d4bfa5dc67d453"
+checksum = "a8725c0a622b374d6cb051d11a0983786448f7785336139c3c94f5aa6bef7e50"
 dependencies = [
  "proc-macro2",
  "pyo3-macros-backend",
@@ -150,9 +182,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros-backend"
-version = "0.22.6"
+version = "0.25.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "36c011a03ba1e50152b4b394b479826cad97e7a21eb52df179cd91ac411cbfbe"
+checksum = "4109984c22491085343c05b0dbc54ddc405c3cf7b4374fc533f5c3313a572ccc"
 dependencies = [
  "heck",
  "proc-macro2",
@@ -176,12 +208,38 @@ version = "5.3.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "69cdb34c158ceb288df11e18b4bd39de994f6657d83847bdffdbd7f346754b0f"
 
+[[package]]
+name = "rayon"
+version = "1.11.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "368f01d005bf8fd9b1206fb6fa653e6c4a81ceb1466406b81792d87c5677a58f"
+dependencies = [
+ "either",
+ "rayon-core",
+]
+
+[[package]]
+name = "rayon-core"
+version = "1.13.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "22e18b0f0062d30d4230b2e85ff77fdfe4326feb054b9783a3460d8435c8ab91"
+dependencies = [
+ "crossbeam-deque",
+ "crossbeam-utils",
+]
+
 [[package]]
 name = "rustversion"
 version = "1.0.22"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b39cdef0fa800fc44525c84ccb54a029961a8215f9619753635a9c0d2538d46d"
 
+[[package]]
+name = "strsim"
+version = "0.11.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7da8b5736845d9f2fcb837ea5d9e2628564b3b043a70948a3f0b778838c5fb4f"
+
 [[package]]
 name = "syn"
 version = "2.0.108"
@@ -195,9 +253,9 @@ dependencies = [
 
 [[package]]
 name = "target-lexicon"
-version = "0.12.16"
+version = "0.13.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "61c41af27dd6d1e27b1b16b489db798443478cef1f06a660c96db617ba5de3b1"
+checksum = "b1dd07eb858a2067e2f3c7155d54e929265c264e6f37efe3ee7a8d1b5a1dd0ba"
 
 [[package]]
 name = "unicode-ident"

{philoch_bib_sdk-0.3.9/philoch_bib_sdk/rust_scorer → philoch_bib_sdk-0.4.0}/Cargo.toml

@@ -1,25 +1,30 @@
 [package]
-name = "rust_scorer"
-version = "0.3.9"
+name = "philoch_bib_sdk"
+version = "0.4.0"
 edition = "2021"
+readme = "README.md"
 
 [lib]
-name = "rust_scorer"
+name = "_rust"
 crate-type = ["cdylib"]
 
 [dependencies]
 pyo3 = "0.25.0"
+ahash = "0.8"
 rayon = "1.11.0"
 strsim = "0.11.1"
 
+[profile.release]
+opt-level = 3
+lto = true
+codegen-units = 1
+
 [lints.clippy]
 all = "warn"
-# Type safety: prevent silent numeric conversions
 cast_possible_truncation = "warn"
 cast_sign_loss = "warn"
 cast_possible_wrap = "warn"
 cast_lossless = "warn"
-# Efficiency
 redundant_clone = "warn"
 
 [lints.rust]

{philoch_bib_sdk-0.3.9 → philoch_bib_sdk-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: philoch-bib-sdk
-Version: 0.3.9
+Version: 0.4.0
 License-File: LICENSE
 Summary: Standard development kit for the Philosophie Bibliography project
 Author-email: Luis Alejandro Bordo García <luis.bordo@philosophie.ch>

philoch_bib_sdk-0.3.9/philoch_bib_sdk/rust_scorer/rust_scorer.pyi → philoch_bib_sdk-0.4.0/philoch_bib_sdk/_rust.pyi

@@ -1,9 +1,56 @@
-"""Type stubs for rust_scorer - high-performance fuzzy matching for BibItems."""
+"""Type stubs for philoch_bib_sdk._rust - Rust extension module.
+
+This module provides high-performance Rust implementations for:
+- Building search indexes for fuzzy matching
+- Batch fuzzy scoring of bibliographic items
+"""
 
 from typing import TypedDict
 
+# === Index Building Types ===
+
+class ItemData(TypedDict):
+    """Input data for a single bibliographic item (for index building)."""
+
+    item_index: int
+    doi: str | None
+    title: str
+    author_surnames: list[str]
+    year: int | None
+    journal_name: str | None
+
+class IndexData:
+    """Output index data structure from build_index_rust."""
+
+    doi_to_index: dict[str, int]
+    trigram_to_indices: dict[str, list[int]]
+    surname_to_indices: dict[str, list[int]]
+    decade_to_indices: dict[int | None, list[int]]
+    journal_to_indices: dict[str, list[int]]
+
+def build_index_rust(items_data: list[ItemData]) -> IndexData:
+    """Build index for fuzzy matching.
+
+    Args:
+        items_data: List of ItemData dicts with bibliographic info
+
+    Returns:
+        IndexData with all indexes built for fast lookup
+    """
+    ...
+
+def hello_rust() -> str:
+    """A simple test function to verify Rust integration works.
+
+    Returns:
+        A greeting string from Rust
+    """
+    ...
+
+# === Scorer Types ===
+
 class BibItemData(TypedDict):
-    """Input data for a single BibItem."""
+    """Input data for a single BibItem (for scoring)."""
 
     index: int
     title: str

{philoch_bib_sdk-0.3.9 → philoch_bib_sdk-0.4.0}/philoch_bib_sdk/logic/functions/fuzzy_matcher.py

@@ -27,11 +27,11 @@ from philoch_bib_sdk.logic.models_staging import PartialScore, ScoreComponent
 
 
 if TYPE_CHECKING:
-    from rust_scorer import BibItemData
+    from philoch_bib_sdk._rust import BibItemData, ItemData
 
 # Try to import Rust scorer for batch processing
 try:
-    import rust_scorer
+    from philoch_bib_sdk import _rust as rust_scorer
 
     _RUST_SCORER_AVAILABLE = True
 except ImportError:
@@ -124,7 +124,7 @@ def _get_decade(date: BibItemDateAttr | str) -> int | None:
     return None
 
 
-def _prepare_items_for_rust(bibitems: Sequence[BibItem]) -> list[dict[str, Any]]:
+def _prepare_items_for_rust(bibitems: Sequence[BibItem]) -> "list[ItemData]":
     """Extract minimal data needed by Rust build_index_rust.
 
     Args:
@@ -134,7 +134,7 @@ def _prepare_items_for_rust(bibitems: Sequence[BibItem]) -> list[dict[str, Any]]
         List of dicts with minimal data for Rust
     """
 
-    items_data = []
+    items_data: list[ItemData] = []
     for i, item in enumerate(bibitems):
         # Extract title string
         title_attr = item.title
@@ -297,7 +297,7 @@ def build_index(bibitems: Sequence[BibItem]) -> BibItemBlockIndex:
     """
     # Try to use Rust implementation
     try:
-        from philoch_bib_sdk._rust import build_index_rust  # type: ignore[import-not-found]
+        from philoch_bib_sdk._rust import build_index_rust
 
         use_rust = True
     except ImportError:

{philoch_bib_sdk-0.3.9 → philoch_bib_sdk-0.4.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "philoch-bib-sdk"
-version = "0.3.9"
+version = "0.4.0"
 description = "Standard development kit for the Philosophie Bibliography project"
 authors = [
     {name = "Luis Alejandro Bordo García", email = "luis.bordo@philosophie.ch"}
@@ -19,6 +19,14 @@ build-backend = "maturin"
 # Note: For development, continue using Poetry commands (poetry install, poetry run, etc.)
 # Poetry will work alongside maturin. Use 'maturin develop' to build the Rust extension.
 # For PyPI releases, use 'maturin build --release'
+  
+dependencies = [                                                                             
+    "aletk>=0.1.6",
+    "attrs>=25.3.0",
+    "polars>=1.32.3",
+    "pydantic>=2.11.9",
+    "cytoolz>=1.0.1",
+]                                                                                            
 
 [tool.poetry.dependencies]
 aletk = "^0.1.6"

{philoch_bib_sdk-0.3.9/philoch_bib_sdk/rust_scorer → philoch_bib_sdk-0.4.0}/src/lib.rs

@@ -1,9 +1,155 @@
 use pyo3::prelude::*;
+use pyo3::types::{PyDict, PyList};
 use rayon::prelude::*;
 use std::cmp::Ordering;
-use std::collections::BinaryHeap;
+use std::collections::{BinaryHeap, HashSet};
 use strsim::jaro_winkler;
 
+/// Input data for a single bibliographic item
+#[derive(Debug, FromPyObject)]
+struct ItemData {
+    #[pyo3(item)]
+    item_index: usize,
+    #[pyo3(item)]
+    doi: Option<String>,
+    #[pyo3(item)]
+    title: String,
+    #[pyo3(item)]
+    author_surnames: Vec<String>,
+    #[pyo3(item)]
+    year: Option<i32>,
+    #[pyo3(item)]
+    journal_name: Option<String>,
+}
+
+/// Output index data structure
+#[pyclass]
+struct IndexData {
+    #[pyo3(get)]
+    doi_to_index: Py<PyDict>,
+    #[pyo3(get)]
+    trigram_to_indices: Py<PyDict>,
+    #[pyo3(get)]
+    surname_to_indices: Py<PyDict>,
+    #[pyo3(get)]
+    decade_to_indices: Py<PyDict>,
+    #[pyo3(get)]
+    journal_to_indices: Py<PyDict>,
+}
+
+/// Extract trigrams from text
+fn extract_trigrams(text: &str) -> HashSet<String> {
+    let normalized = text.to_lowercase();
+    let chars: Vec<char> = normalized.chars().collect();
+
+    if chars.len() < 3 {
+        return HashSet::new();
+    }
+
+    let mut trigrams = HashSet::new();
+    for i in 0..=chars.len() - 3 {
+        let trigram: String = chars[i..i + 3].iter().collect();
+        trigrams.insert(trigram);
+    }
+
+    trigrams
+}
+
+/// Calculate decade from year
+fn get_decade(year: Option<i32>) -> Option<i32> {
+    year.map(|y| (y / 10) * 10)
+}
+
+/// Build index for fuzzy matching
+#[pyfunction]
+fn build_index_rust(py: Python, items_data: Vec<ItemData>) -> PyResult<IndexData> {
+    use ahash::AHashMap;
+
+    // Pre-allocate with capacity hints
+    let capacity = items_data.len();
+    let mut doi_map: AHashMap<String, usize> = AHashMap::with_capacity(capacity);
+    let mut trigram_map: AHashMap<String, Vec<usize>> = AHashMap::new();
+    let mut surname_map: AHashMap<String, Vec<usize>> = AHashMap::new();
+    let mut decade_map: AHashMap<Option<i32>, Vec<usize>> = AHashMap::new();
+    let mut journal_map: AHashMap<String, Vec<usize>> = AHashMap::new();
+
+    // Single pass over all items
+    for item in items_data {
+        let idx = item.item_index;
+
+        // DOI index
+        if let Some(doi) = item.doi {
+            doi_map.insert(doi, idx);
+        }
+
+        // Title trigram index
+        let trigrams = extract_trigrams(&item.title);
+        for trigram in trigrams {
+            trigram_map.entry(trigram).or_default().push(idx);
+        }
+
+        // Author surname index
+        for surname in item.author_surnames {
+            let normalized = surname.to_lowercase().trim().to_string();
+            if !normalized.is_empty() {
+                surname_map.entry(normalized).or_default().push(idx);
+            }
+        }
+
+        // Year decade index
+        let decade = get_decade(item.year);
+        decade_map.entry(decade).or_default().push(idx);
+
+        // Journal index
+        if let Some(journal) = item.journal_name {
+            let normalized = journal.to_lowercase().trim().to_string();
+            if !normalized.is_empty() {
+                journal_map.entry(normalized).or_default().push(idx);
+            }
+        }
+    }
+
+    // Convert to Python dicts
+    let doi_dict = PyDict::new(py);
+    for (k, v) in doi_map {
+        doi_dict.set_item(k, v)?;
+    }
+
+    let trigram_dict = PyDict::new(py);
+    for (k, v) in trigram_map {
+        let py_list = PyList::new(py, &v)?;
+        trigram_dict.set_item(k, py_list)?;
+    }
+
+    let surname_dict = PyDict::new(py);
+    for (k, v) in surname_map {
+        let py_list = PyList::new(py, &v)?;
+        surname_dict.set_item(k, py_list)?;
+    }
+
+    let decade_dict = PyDict::new(py);
+    for (k, v) in decade_map {
+        let py_list = PyList::new(py, &v)?;
+        decade_dict.set_item(k, py_list)?;
+    }
+
+    let journal_dict = PyDict::new(py);
+    for (k, v) in journal_map {
+        let py_list = PyList::new(py, &v)?;
+        journal_dict.set_item(k, py_list)?;
+    }
+
+    Ok(IndexData {
+        doi_to_index: doi_dict.into(),
+        trigram_to_indices: trigram_dict.into(),
+        surname_to_indices: surname_dict.into(),
+        decade_to_indices: decade_dict.into(),
+        journal_to_indices: journal_dict.into(),
+    })
+}
+
+// === SCORER FUNCTIONALITY (merged from rust_scorer) ===
+
 /// Normalize text: lowercase and collapse whitespace
 fn normalize(s: &str) -> String {
     s.to_lowercase()
@@ -41,7 +187,7 @@ fn token_sort_ratio(s1: &str, s2: &str) -> f64 {
     token_sort_ratio_f64(s1, s2)
 }
 
-/// Input data for a single BibItem (simplified for Rust processing)
+/// Input data for a single BibItem (for scoring)
 #[derive(Clone, Debug, FromPyObject)]
 #[pyo3(from_item_all)]
 struct BibItemData {
@@ -306,9 +452,21 @@ fn score_batch(
         .collect()
 }
 
-/// A Python module implemented in Rust for fast fuzzy matching.
+// === END SCORER FUNCTIONALITY ===
+
+/// A simple test function to verify Rust integration works
+#[pyfunction]
+fn hello_rust() -> PyResult<String> {
+    Ok("Hello from Rust!".to_string())
+}
+
+/// A Python module implemented in Rust.
 #[pymodule]
-fn rust_scorer(m: &Bound<'_, PyModule>) -> PyResult<()> {
+fn _rust(m: &Bound<'_, PyModule>) -> PyResult<()> {
+    m.add_function(wrap_pyfunction!(hello_rust, m)?)?;
+    m.add_function(wrap_pyfunction!(build_index_rust, m)?)?;
+    m.add_class::<IndexData>()?;
+    // Scorer functions (merged from rust_scorer)
     m.add_function(wrap_pyfunction!(token_sort_ratio, m)?)?;
     m.add_function(wrap_pyfunction!(score_batch, m)?)?;
     Ok(())
@@ -318,6 +476,33 @@ fn rust_scorer(m: &Bound<'_, PyModule>) -> PyResult<()> {
 mod tests {
     use super::*;
 
+    #[test]
+    fn test_extract_trigrams() {
+        let text = "hello";
+        let trigrams = extract_trigrams(text);
+
+        assert_eq!(trigrams.len(), 3);
+        assert!(trigrams.contains("hel"));
+        assert!(trigrams.contains("ell"));
+        assert!(trigrams.contains("llo"));
+    }
+
+    #[test]
+    fn test_extract_trigrams_short() {
+        let text = "hi";
+        let trigrams = extract_trigrams(text);
+        assert_eq!(trigrams.len(), 0);
+    }
+
+    #[test]
+    fn test_get_decade() {
+        assert_eq!(get_decade(Some(1995)), Some(1990));
+        assert_eq!(get_decade(Some(2000)), Some(2000));
+        assert_eq!(get_decade(Some(2025)), Some(2020));
+        assert_eq!(get_decade(None), None);
+    }
+
+    // Scorer tests (merged from rust_scorer)
     #[test]
     fn test_token_sort_ratio_identical() {
         let score = token_sort_ratio("hello world", "hello world");

{philoch_bib_sdk-0.3.9 → philoch_bib_sdk-0.4.0}/tests/logic/functions/test_fuzzy_matcher.py

@@ -5,7 +5,7 @@ from typing import TYPE_CHECKING
 from philoch_bib_sdk.logic.default_models import default_bib_item
 
 if TYPE_CHECKING:
-    from rust_scorer import BibItemData
+    from philoch_bib_sdk._rust import BibItemData
 from philoch_bib_sdk.logic.functions.fuzzy_matcher import (
     BibItemBlockIndex,
     build_index,
@@ -500,7 +500,7 @@ def test_rust_scorer_available() -> None:
 
         pytest.skip("Rust scorer not available")
 
-    import rust_scorer
+    from philoch_bib_sdk import _rust as rust_scorer
 
     # Test basic function exists and works
     score = rust_scorer.token_sort_ratio("hello world", "world hello")
@@ -516,7 +516,7 @@ def test_rust_batch_scorer_basic() -> None:
 
         pytest.skip("Rust scorer not available")
 
-    import rust_scorer
+    from philoch_bib_sdk import _rust as rust_scorer
 
     subjects: list[BibItemData] = [
         {

philoch_bib_sdk-0.3.9/Cargo.toml

@@ -1,18 +0,0 @@
-[package]
-name = "philoch_bib_sdk"
-version = "0.1.6"
-edition = "2021"
-readme = "README.md"
-
-[lib]
-name = "_rust"
-crate-type = ["cdylib"]
-
-[dependencies]
-pyo3 = { version = "0.22", features = ["extension-module"] }
-ahash = "0.8"
-
-[profile.release]
-opt-level = 3
-lto = true
-codegen-units = 1