rapid-textrank 0.1.0__tar.gz → 0.1.1__tar.gz

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/Cargo.lock

@@ -579,7 +579,7 @@ dependencies = [
 
 [[package]]
 name = "rapid_textrank"
-version = "0.1.0"
+version = "0.1.1"
 dependencies = [
  "approx",
  "criterion",

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "rapid_textrank"
-version = "0.1.0"
+version = "0.1.1"
 edition = "2021"
 authors = ["TextRanker Contributors"]
 description = "High-performance TextRank implementation with Python bindings"

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rapid_textrank
-Version: 0.1.0
+Version: 0.1.1
 Classifier: Development Status :: 4 - Beta
 Classifier: Intended Audience :: Developers
 Classifier: Intended Audience :: Science/Research
@@ -217,12 +217,16 @@ config = TextRankConfig(
     damping=0.85,              # PageRank damping factor (0-1)
     max_iterations=100,        # Maximum PageRank iterations
     convergence_threshold=1e-6,# Convergence threshold
-    window_size=4,             # Co-occurrence window size
+    window_size=3,             # Co-occurrence window size
     top_n=10,                  # Number of results
     min_phrase_length=1,       # Minimum words in a phrase
     max_phrase_length=4,       # Maximum words in a phrase
     score_aggregation="sum",   # How to combine word scores: "sum", "mean", "max", "rms"
-    language="en"              # Language for stopwords
+    language="en",             # Language for stopwords
+    include_pos=["NOUN","ADJ","PROPN","VERB"],  # POS tags to include in the graph
+    use_pos_in_nodes=True,     # If True, graph nodes are lemma+POS
+    phrase_grouping="scrubbed_text",   # "lemma" or "scrubbed_text"
+    stopwords=["custom", "terms"]  # Additional stopwords (extends built-in list)
 )
 
 extractor = BaseTextRank(config=config)

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/README.md

@@ -182,12 +182,16 @@ config = TextRankConfig(
     damping=0.85,              # PageRank damping factor (0-1)
     max_iterations=100,        # Maximum PageRank iterations
     convergence_threshold=1e-6,# Convergence threshold
-    window_size=4,             # Co-occurrence window size
+    window_size=3,             # Co-occurrence window size
     top_n=10,                  # Number of results
     min_phrase_length=1,       # Minimum words in a phrase
     max_phrase_length=4,       # Maximum words in a phrase
     score_aggregation="sum",   # How to combine word scores: "sum", "mean", "max", "rms"
-    language="en"              # Language for stopwords
+    language="en",             # Language for stopwords
+    include_pos=["NOUN","ADJ","PROPN","VERB"],  # POS tags to include in the graph
+    use_pos_in_nodes=True,     # If True, graph nodes are lemma+POS
+    phrase_grouping="scrubbed_text",   # "lemma" or "scrubbed_text"
+    stopwords=["custom", "terms"]  # Additional stopwords (extends built-in list)
 )
 
 extractor = BaseTextRank(config=config)

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/notebooks/01_quickstart.ipynb

@@ -193,17 +193,11 @@
    "source": [
     "from rapid_textrank import TextRankConfig, BaseTextRank\n",
     "\n",
-    "# Create a custom configuration\n",
+    "# Create a custom configuration (only overriding a few defaults)\n",
     "config = TextRankConfig(\n",
-    "    damping=0.85,              # PageRank damping factor (0-1)\n",
-    "    max_iterations=100,        # Maximum PageRank iterations\n",
-    "    convergence_threshold=1e-6,# Stop when scores change less than this\n",
-    "    window_size=4,             # Co-occurrence window size\n",
-    "    top_n=10,                  # Number of results\n",
-    "    min_phrase_length=1,       # Minimum words in a phrase\n",
-    "    max_phrase_length=4,       # Maximum words in a phrase\n",
-    "    score_aggregation=\"sum\",   # How to combine word scores: \"sum\", \"mean\", \"max\", \"rms\"\n",
-    "    language=\"en\"              # Language for stopwords\n",
+    "    top_n=10,\n",
+    "    score_aggregation=\"sum\",\n",
+    "    language=\"en\",\n",
     ")\n",
     "\n",
     "# Create an extractor with the config\n",
@@ -214,9 +208,10 @@
     "\n",
     "print(f\"Converged: {result.converged}\")\n",
     "print(f\"Iterations: {result.iterations}\")\n",
-    "print(f\"\\nTop phrases:\")\n",
+    "print(f\"\n",
+    "Top phrases:\")\n",
     "for p in result.phrases[:5]:\n",
-    "    print(f\"  {p.rank}. {p.text}: {p.score:.4f}\")"
+    "    print(f\"  {p.rank}. {p.text}: {p.score:.4f}\")\n"
    ]
   },
   {
@@ -317,7 +312,7 @@
      "output_type": "stream",
      "text": [
       "German keywords:\n",
-      "  1. Teilgebiet der künstlichen Intelligenz: 0.1860\n",
+      "  1. Teilgebiet der k\u00fcnstlichen Intelligenz: 0.1860\n",
       "  2. aus Erfahrung zu lernen: 0.1768\n",
       "  3. Netze mit vielen Schichten: 0.1184\n"
      ]
@@ -326,9 +321,9 @@
    "source": [
     "# German example\n",
     "german_text = \"\"\"\n",
-    "Maschinelles Lernen ist ein Teilgebiet der künstlichen Intelligenz.\n",
+    "Maschinelles Lernen ist ein Teilgebiet der k\u00fcnstlichen Intelligenz.\n",
     "Deep Learning verwendet neuronale Netze mit vielen Schichten.\n",
-    "Diese Technologie ermöglicht es Computern, aus Erfahrung zu lernen.\n",
+    "Diese Technologie erm\u00f6glicht es Computern, aus Erfahrung zu lernen.\n",
     "\"\"\"\n",
     "\n",
     "keywords_de = extract_keywords(german_text, top_n=5, language=\"de\")\n",
@@ -350,7 +345,7 @@
      "text": [
       "French keywords:\n",
       "  1. branche de l'intelligence artificielle: 0.1906\n",
-      "  2. l'analyse de données complexes: 0.1764\n",
+      "  2. l'analyse de donn\u00e9es complexes: 0.1764\n",
       "  3. de nombreux secteurs industriels: 0.1250\n"
      ]
     }
@@ -359,7 +354,7 @@
     "# French example\n",
     "french_text = \"\"\"\n",
     "L'apprentissage automatique est une branche de l'intelligence artificielle.\n",
-    "Les réseaux de neurones profonds permettent l'analyse de données complexes.\n",
+    "Les r\u00e9seaux de neurones profonds permettent l'analyse de donn\u00e9es complexes.\n",
     "Ces technologies transforment de nombreux secteurs industriels.\n",
     "\"\"\"\n",
     "\n",
@@ -395,10 +390,10 @@
       "Note: you may need to restart the kernel to use updated packages.\n",
       "Collecting en-core-web-sm==3.8.0\n",
       "  Downloading https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.8.0/en_core_web_sm-3.8.0-py3-none-any.whl (12.8 MB)\n",
-      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m12.8/12.8 MB\u001b[0m \u001b[31m37.8 MB/s\u001b[0m  \u001b[33m0:00:00\u001b[0meta \u001b[36m0:00:01\u001b[0m\n",
+      "\u001b[2K     \u001b[90m\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u001b[0m \u001b[32m12.8/12.8 MB\u001b[0m \u001b[31m37.8 MB/s\u001b[0m  \u001b[33m0:00:00\u001b[0meta \u001b[36m0:00:01\u001b[0m\n",
       "\u001b[?25hInstalling collected packages: en-core-web-sm\n",
       "Successfully installed en-core-web-sm-3.8.0\n",
-      "\u001b[38;5;2m✔ Download and installation successful\u001b[0m\n",
+      "\u001b[38;5;2m\u2714 Download and installation successful\u001b[0m\n",
       "You can now load the package via spacy.load('en_core_web_sm')\n"
      ]
     }
@@ -484,4 +479,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/notebooks/02_algorithm_variants.ipynb

@@ -118,7 +118,7 @@
     "\n",
     "Based on [Florescu & Caragea (2017)](https://aclanthology.org/P17-1102/), PositionRank weights words by their position in the document.\n",
     "\n",
-    "**Key insight:** In many documents (papers, news articles, reports), important terms appear early—in titles, abstracts, or introductory paragraphs.\n",
+    "**Key insight:** In many documents (papers, news articles, reports), important terms appear early\u2014in titles, abstracts, or introductory paragraphs.\n",
     "\n",
     "**How it differs from BaseTextRank:**\n",
     "- Words appearing early get higher initial importance\n",
@@ -555,7 +555,7 @@
     "    ],\n",
     "    \"config\": {\n",
     "        \"top_n\": 5,\n",
-    "        \"window_size\": 4,\n",
+    "        \"window_size\": 3,\n",
     "        \"damping\": 0.85\n",
     "    }\n",
     "}\n",
@@ -565,7 +565,7 @@
     "\n",
     "print(\"Single document result:\")\n",
     "for phrase in result[\"phrases\"]:\n",
-    "    print(f\"  {phrase['text']}: {phrase['score']:.4f}\")"
+    "    print(f\"  {phrase['text']}: {phrase['score']:.4f}\")\n"
    ]
   },
   {
@@ -639,27 +639,27 @@
     "\n",
     "```\n",
     "                                START\n",
-    "                                  │\n",
-    "                                  ▼\n",
-    "                    ┌─────────────────────────┐\n",
-    "                    │ Do you have specific    │\n",
-    "                    │ topics to focus on?     │\n",
-    "                    └─────────────────────────┘\n",
-    "                         │              │\n",
+    "                                  \u2502\n",
+    "                                  \u25bc\n",
+    "                    \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n",
+    "                    \u2502 Do you have specific    \u2502\n",
+    "                    \u2502 topics to focus on?     \u2502\n",
+    "                    \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n",
+    "                         \u2502              \u2502\n",
     "                        YES             NO\n",
-    "                         │              │\n",
-    "                         ▼              ▼\n",
-    "               ┌──────────────┐  ┌─────────────────────────┐\n",
-    "               │ BiasedTextRank│  │ Is key info at the     │\n",
-    "               │              │  │ beginning of the doc?   │\n",
-    "               └──────────────┘  └─────────────────────────┘\n",
-    "                                       │              │\n",
+    "                         \u2502              \u2502\n",
+    "                         \u25bc              \u25bc\n",
+    "               \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510  \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n",
+    "               \u2502 BiasedTextRank\u2502  \u2502 Is key info at the     \u2502\n",
+    "               \u2502              \u2502  \u2502 beginning of the doc?   \u2502\n",
+    "               \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518  \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n",
+    "                                       \u2502              \u2502\n",
     "                                      YES             NO\n",
-    "                                       │              │\n",
-    "                                       ▼              ▼\n",
-    "                              ┌──────────────┐ ┌──────────────┐\n",
-    "                              │ PositionRank │ │ BaseTextRank │\n",
-    "                              └──────────────┘ └──────────────┘\n",
+    "                                       \u2502              \u2502\n",
+    "                                       \u25bc              \u25bc\n",
+    "                              \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n",
+    "                              \u2502 PositionRank \u2502 \u2502 BaseTextRank \u2502\n",
+    "                              \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n",
     "```\n",
     "\n",
     "### Recommendations by Document Type\n",
@@ -707,4 +707,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/notebooks/03_explain_algorithm.ipynb

@@ -179,7 +179,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "def build_cooccurrence_graph(tokens, window_size=4):\n",
+    "def build_cooccurrence_graph(tokens, window_size=3):\n",
     "    \"\"\"Build a co-occurrence graph from tokens.\"\"\"\n",
     "    G = nx.Graph()\n",
     "    \n",
@@ -205,11 +205,11 @@
     "    return G\n",
     "\n",
     "# Build graph with default window size\n",
-    "G = build_cooccurrence_graph(tokens, window_size=4)\n",
+    "G = build_cooccurrence_graph(tokens, window_size=3)\n",
     "\n",
     "print(f\"Graph Statistics:\")\n",
     "print(f\"  Nodes: {G.number_of_nodes()}\")\n",
-    "print(f\"  Edges: {G.number_of_edges()}\")"
+    "print(f\"  Edges: {G.number_of_edges()}\")\n"
    ]
   },
   {
@@ -241,10 +241,10 @@
     "edge_labels = {(u, v): G[u][v]['weight'] for u, v in G.edges()}\n",
     "nx.draw_networkx_edge_labels(G, pos, edge_labels, font_size=8, ax=ax)\n",
     "\n",
-    "ax.set_title(\"Co-occurrence Graph (window_size=4)\", fontsize=14, fontweight='bold')\n",
+    "ax.set_title(\"Co-occurrence Graph (window_size=3)\", fontsize=14, fontweight='bold')\n",
     "ax.axis('off')\n",
     "plt.tight_layout()\n",
-    "plt.show()"
+    "plt.show()\n"
    ]
   },
   {
@@ -345,13 +345,13 @@
     "    return scores, history\n",
     "\n",
     "# Run PageRank\n",
-    "G = build_cooccurrence_graph(tokens, window_size=4)\n",
+    "G = build_cooccurrence_graph(tokens, window_size=3)\n",
     "scores, history = pagerank_with_history(G)\n",
     "\n",
     "print(f\"PageRank converged in {len(history)-1} iterations\")\n",
     "print(f\"\\nFinal scores (sorted by importance):\")\n",
     "for word, score in sorted(scores.items(), key=lambda x: -x[1]):\n",
-    "    print(f\"  {word:<15} {score:.4f}\")"
+    "    print(f\"  {word:<15} {score:.4f}\")\n"
    ]
   },
   {
@@ -391,7 +391,7 @@
     "# Visualize graph with node size proportional to score\n",
     "fig, ax = plt.subplots(figsize=(12, 8))\n",
     "\n",
-    "G = build_cooccurrence_graph(tokens, window_size=4)\n",
+    "G = build_cooccurrence_graph(tokens, window_size=3)\n",
     "pos = nx.spring_layout(G, k=2, iterations=50, seed=42)\n",
     "\n",
     "# Node sizes based on PageRank scores\n",
@@ -414,7 +414,7 @@
     "             fontsize=14, fontweight='bold')\n",
     "ax.axis('off')\n",
     "plt.tight_layout()\n",
-    "plt.show()"
+    "plt.show()\n"
    ]
   },
   {

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/notebooks/04_benchmarks.ipynb

@@ -27,7 +27,7 @@
      "output_type": "stream",
      "text": [
       "Note: you may need to restart the kernel to use updated packages.\n",
-      "\u001b[38;5;2m✔ Download and installation successful\u001b[0m\n",
+      "\u001b[38;5;2m\u2714 Download and installation successful\u001b[0m\n",
       "You can now load the package via spacy.load('en_core_web_sm')\n"
      ]
     }
@@ -240,20 +240,20 @@
       "\n",
       "SMALL TEXT (~16 words)\n",
       "--------------------------------------------------\n",
-      "rapid_textrank:      4.44 ms (±1.38)\n",
-      "pytextrank:          7.54 ms (±1.00)\n",
+      "rapid_textrank:      4.44 ms (\u00b11.38)\n",
+      "pytextrank:          7.54 ms (\u00b11.00)\n",
       "Speedup:              1.7x faster\n",
       "\n",
       "MEDIUM TEXT (~100 words)\n",
       "--------------------------------------------------\n",
-      "rapid_textrank:      3.08 ms (±0.32)\n",
-      "pytextrank:         66.14 ms (±104.73)\n",
+      "rapid_textrank:      3.08 ms (\u00b10.32)\n",
+      "pytextrank:         66.14 ms (\u00b1104.73)\n",
       "Speedup:             21.4x faster\n",
       "\n",
       "LARGE TEXT (~660 words)\n",
       "--------------------------------------------------\n",
-      "rapid_textrank:      4.12 ms (±0.83)\n",
-      "pytextrank:        184.31 ms (±26.20)\n",
+      "rapid_textrank:      4.12 ms (\u00b10.83)\n",
+      "pytextrank:        184.31 ms (\u00b126.20)\n",
       "Speedup:             44.7x faster\n"
      ]
     }
@@ -272,11 +272,11 @@
     "\n",
     "    # Benchmark rapid_textrank\n",
     "    rust_results = benchmark_rapid_textrank(text)\n",
-    "    print(f\"rapid_textrank:  {rust_results['mean']:>8.2f} ms (±{rust_results['std']:.2f})\")\n",
+    "    print(f\"rapid_textrank:  {rust_results['mean']:>8.2f} ms (\u00b1{rust_results['std']:.2f})\")\n",
     "\n",
     "    # Benchmark pytextrank\n",
     "    py_results = benchmark_pytextrank(text)\n",
-    "    print(f\"pytextrank:      {py_results['mean']:>8.2f} ms (±{py_results['std']:.2f})\")\n",
+    "    print(f\"pytextrank:      {py_results['mean']:>8.2f} ms (\u00b1{py_results['std']:.2f})\")\n",
     "\n",
     "    speedup = py_results['mean'] / rust_results['mean']\n",
     "    print(f\"Speedup:         {speedup:>8.1f}x faster\")\n",
@@ -496,20 +496,20 @@
       "\n",
       "SMALL TEXT (21 tokens)\n",
       "--------------------------------------------------\n",
-      "rapid_textrank (JSON API):     0.037 ms (±0.010)\n",
-      "pytextrank (extraction):       1.596 ms (±0.582)\n",
+      "rapid_textrank (JSON API):     0.037 ms (\u00b10.010)\n",
+      "pytextrank (extraction):       1.596 ms (\u00b10.582)\n",
       "Speedup:                        42.9x faster\n",
       "\n",
       "MEDIUM TEXT (138 tokens)\n",
       "--------------------------------------------------\n",
-      "rapid_textrank (JSON API):     0.231 ms (±0.101)\n",
-      "pytextrank (extraction):       2.881 ms (±0.472)\n",
+      "rapid_textrank (JSON API):     0.231 ms (\u00b10.101)\n",
+      "pytextrank (extraction):       2.881 ms (\u00b10.472)\n",
       "Speedup:                        12.5x faster\n",
       "\n",
       "LARGE TEXT (838 tokens)\n",
       "--------------------------------------------------\n",
-      "rapid_textrank (JSON API):     0.869 ms (±0.068)\n",
-      "pytextrank (extraction):       9.171 ms (±0.891)\n",
+      "rapid_textrank (JSON API):     0.869 ms (\u00b10.068)\n",
+      "pytextrank (extraction):       9.171 ms (\u00b10.891)\n",
       "Speedup:                        10.5x faster\n"
      ]
     }
@@ -529,10 +529,10 @@
     "    print(\"-\" * 50)\n",
     "    \n",
     "    rapid_res = benchmark_rapid_extraction_only(tokens)\n",
-    "    print(f\"rapid_textrank (JSON API):  {rapid_res['mean']:>8.3f} ms (±{rapid_res['std']:.3f})\")\n",
+    "    print(f\"rapid_textrank (JSON API):  {rapid_res['mean']:>8.3f} ms (\u00b1{rapid_res['std']:.3f})\")\n",
     "    \n",
     "    py_res = benchmark_pytextrank_extraction_only(doc)\n",
-    "    print(f\"pytextrank (extraction):    {py_res['mean']:>8.3f} ms (±{py_res['std']:.3f})\")\n",
+    "    print(f\"pytextrank (extraction):    {py_res['mean']:>8.3f} ms (\u00b1{py_res['std']:.3f})\")\n",
     "    \n",
     "    speedup = py_res['mean'] / rapid_res['mean'] if rapid_res['mean'] > 0 else float('inf')\n",
     "    print(f\"Speedup:                    {speedup:>8.1f}x faster\")\n",
@@ -647,9 +647,9 @@
      "text": [
       "Benchmarking batch processing (100 documents)...\n",
       "============================================================\n",
-      "rapid_textrank (batch JSON):      18.50 ms (±0.37)\n",
+      "rapid_textrank (batch JSON):      18.50 ms (\u00b10.37)\n",
       "  Per document:                   0.185 ms\n",
-      "pytextrank (sequential):        3431.75 ms (±849.32)\n",
+      "pytextrank (sequential):        3431.75 ms (\u00b1849.32)\n",
       "  Per document:                  34.317 ms\n",
       "\n",
       "Speedup: 185.5x faster\n"
@@ -697,11 +697,11 @@
     "print(\"=\" * 60)\n",
     "\n",
     "rapid_mean, rapid_std = benchmark_rapid_batch(batch_tokens)\n",
-    "print(f\"rapid_textrank (batch JSON): {rapid_mean:>10.2f} ms (±{rapid_std:.2f})\")\n",
+    "print(f\"rapid_textrank (batch JSON): {rapid_mean:>10.2f} ms (\u00b1{rapid_std:.2f})\")\n",
     "print(f\"  Per document:              {rapid_mean/num_docs:>10.3f} ms\")\n",
     "\n",
     "py_mean, py_std = benchmark_pytextrank_batch(batch_texts)\n",
-    "print(f\"pytextrank (sequential):     {py_mean:>10.2f} ms (±{py_std:.2f})\")\n",
+    "print(f\"pytextrank (sequential):     {py_mean:>10.2f} ms (\u00b1{py_std:.2f})\")\n",
     "print(f\"  Per document:              {py_mean/num_docs:>10.3f} ms\")\n",
     "\n",
     "speedup = py_mean / rapid_mean\n",
@@ -978,4 +978,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "maturin"
 
 [project]
 name = "rapid_textrank"
-version = "0.1.0"
+version = "0.1.1"
 description = "High-performance TextRank implementation with Python bindings"
 readme = { file = "README.md", content-type = "text/markdown" }
 license = "MIT"

{rapid_textrank-0.1.0 → rapid_textrank-0.1.1}/python/rapid_textrank/spacy_component.py

@@ -79,11 +79,16 @@ if SPACY_AVAILABLE:
             "damping": 0.85,
             "max_iterations": 100,
             "convergence_threshold": 1e-6,
-            "window_size": 4,
+            "window_size": 3,
             "top_n": 10,
             "min_phrase_length": 1,
             "max_phrase_length": 4,
             "score_aggregation": "sum",
+            "include_pos": ["ADJ", "NOUN", "PROPN", "VERB"],
+            "use_pos_in_nodes": True,
+            "phrase_grouping": "scrubbed_text",
+            "language": "en",
+            "stopwords": None,
         },
     )
     def create_rapid_textrank(
@@ -97,6 +102,11 @@ if SPACY_AVAILABLE:
         min_phrase_length: int,
         max_phrase_length: int,
         score_aggregation: str,
+        include_pos: Optional[List[str]],
+        use_pos_in_nodes: bool,
+        phrase_grouping: str,
+        language: str,
+        stopwords: Optional[List[str]],
     ):
         """Create a RustTextRank pipeline component."""
         return RustTextRank(
@@ -110,6 +120,11 @@ if SPACY_AVAILABLE:
             min_phrase_length=min_phrase_length,
             max_phrase_length=max_phrase_length,
             score_aggregation=score_aggregation,
+            include_pos=include_pos,
+            use_pos_in_nodes=use_pos_in_nodes,
+            phrase_grouping=phrase_grouping,
+            language=language,
+            stopwords=stopwords,
         )
 
     class RustTextRank:
@@ -135,11 +150,16 @@ if SPACY_AVAILABLE:
             damping: float = 0.85,
             max_iterations: int = 100,
             convergence_threshold: float = 1e-6,
-            window_size: int = 4,
+            window_size: int = 3,
             top_n: int = 10,
             min_phrase_length: int = 1,
             max_phrase_length: int = 4,
             score_aggregation: str = "sum",
+            include_pos: Optional[List[str]] = None,
+            use_pos_in_nodes: bool = True,
+            phrase_grouping: str = "scrubbed_text",
+            language: str = "en",
+            stopwords: Optional[List[str]] = None,
         ):
             self.nlp = nlp
             self.name = name
@@ -152,7 +172,14 @@ if SPACY_AVAILABLE:
                 "min_phrase_length": min_phrase_length,
                 "max_phrase_length": max_phrase_length,
                 "score_aggregation": score_aggregation,
+                "use_pos_in_nodes": use_pos_in_nodes,
+                "phrase_grouping": phrase_grouping,
+                "language": language,
             }
+            if include_pos is not None:
+                self.config["include_pos"] = include_pos
+            if stopwords is not None:
+                self.config["stopwords"] = stopwords
 
             # Register custom extensions
             if not Doc.has_extension("phrases"):