Create index

crawl.py

@@ -4,10 +4,11 @@ Crawl the web
 import gzip
 import hashlib
 import os
+import sys
+from traceback import print_tb, print_exc
 
 import pandas as pd
 import requests
-import justext
 
 from paths import DATA_DIR, HN_TOP_PATH, CRAWL_PREFIX
 
@@ -16,15 +17,23 @@ def crawl():
     data = pd.read_csv(HN_TOP_PATH)
 
     for url in data['url']:
-        print("Fetching", url)
-        html = fetch(url)
         filename = hashlib.md5(url.encode('utf8')).hexdigest()
         path = os.path.join(DATA_DIR, f"{CRAWL_PREFIX}{filename}.html.gz")
         if os.path.isfile(path):
-            print("Path already exists, skipping")
+            print("Path already exists, skipping", url)
+            continue
 
-        with gzip.open(path, 'w') as output:
-            output.write(html.encode('utf8'))
+        print("Fetching", url)
+        try:
+            html = fetch(url)
+        except Exception:
+            print_exc(file=sys.stderr)
+            print("Unable to fetch", url)
+            continue
+
+        with gzip.open(path, 'wt') as output:
+            output.write(url + '\n')
+            output.write(html)
 
 
 def fetch(url):

index.py

@@ -2,12 +2,16 @@
 Create a search index
 """
 import gzip
+import sqlite3
 from glob import glob
 
+import bs4
 import justext
 from spacy.lang.en import English
 
-from paths import CRAWL_GLOB
+from paths import CRAWL_GLOB, INDEX_PATH
+
+NUM_INITIAL_TOKENS = 50
 
 
 def is_content_token(nlp, token):
@@ -17,7 +21,8 @@ def is_content_token(nlp, token):
 
 def tokenize(nlp, cleaned_text):
     tokens = nlp.tokenizer(cleaned_text)
-    content_tokens = [token for token in tokens if is_content_token(nlp, token)]
+    content_tokens = [token for token in tokens[:NUM_INITIAL_TOKENS]
+                      if is_content_token(nlp, token)]
     lowered = {nlp.vocab[token.orth].text.lower() for token in content_tokens}
     return lowered
 
@@ -29,15 +34,79 @@ def clean(content):
     return cleaned_text
 
 
+def index(tokens, url, title):
+    with sqlite3.connect(INDEX_PATH) as con:
+        con.execute("""
+            INSERT INTO pages (url, title)
+            VALUES (?, ?)
+        """, (url, title))
+
+        result = con.execute("""
+            SELECT last_insert_rowid()
+        """)
+        page_id = result.fetchone()[0]
+        print("Created page with id", page_id)
+
+        con.executemany("""
+            INSERT INTO terms (term, page_id)
+            VALUES (?, ?)
+        """, [(term, page_id) for term in tokens])
+
+
+def create_if_not_exists():
+    con = sqlite3.connect(INDEX_PATH)
+    con.execute("""
+    CREATE TABLE IF NOT EXISTS pages (
+      id INTEGER PRIMARY KEY,
+      url TEXT UNIQUE,
+      title TEXT
+    )
+    """)
+
+    con.execute("""
+    CREATE TABLE IF NOT EXISTS terms (
+      term TEXT,
+      page_id INTEGER 
+    )
+    """)
+
+    con.execute("""
+    CREATE INDEX IF NOT EXISTS term_index ON terms (term)
+    """)
+
+
+def page_indexed(url):
+    con = sqlite3.connect(INDEX_PATH)
+    result = con.execute("""
+        SELECT EXISTS(SELECT 1 FROM pages WHERE url=?)
+    """, (url,))
+    value = result.fetchone()[0]
+    return value == 1
+
+
 def run():
+    create_if_not_exists()
     nlp = English()
     for path in glob(CRAWL_GLOB):
-        with gzip.open(path) as html_file:
-            content = html_file.read().decode("utf8")
+        print("Path", path)
+        with gzip.open(path, 'rt') as html_file:
+            url = html_file.readline().strip()
+            content = html_file.read()
+
+        if page_indexed(url):
+            print("Page exists, skipping", url)
+            continue
+
         cleaned_text = clean(content)
+        try:
+            title = bs4.BeautifulSoup(content, features="lxml").find('title').string
+        except AttributeError:
+            title = cleaned_text[:80]
         tokens = tokenize(nlp, cleaned_text)
+        print("URL", url)
         print("Tokens", tokens)
-        break
+        print("Title", title)
+        index(tokens, url, title)
 
 
 if __name__ == '__main__':

paths.py

@@ -5,3 +5,4 @@ DATA_DIR = os.path.join(HOME, 'data', 'tinysearch')
 HN_TOP_PATH = os.path.join(DATA_DIR, 'hn-top.csv')
 CRAWL_PREFIX = 'crawl_'
 CRAWL_GLOB = os.path.join(DATA_DIR, f"{CRAWL_PREFIX}*")
+INDEX_PATH = os.path.join(DATA_DIR, 'index.sqlite3')