files

hw/README.txt

+The assignment instructions and starter code (if needed) are posted here.
+
+Please place your solutions in the submission directory then run:
+
+$ ./turnInHw.sh hw1
+

hw/hw1/starter-code/createRawSchema.sql

+create table Pub (k text, p text);
+create table Field (k text, i text, p text, v text);
+copy Pub from 'pubFile.txt';
+copy Field from 'fieldFile.txt';

hw/hw1/starter-code/pythonpsql.py

+#!/usr/bin/python
+import psycopg2
+
+def main():
+	try:
+	    conn = psycopg2.connect("dbname='dblp' user='<YOUR USER NAME>' host='localhost' password=''")
+	except psycopg2.Error, e:
+	    print "I am unable to connect to the database"
+
+	cur = conn.cursor()
+
+	cur.execute("SELECT * FROM author LIMIT 10")
+
+	rows = cur.fetchall()
+
+	print "Showing first 10 results:\n"
+
+	for row in rows:
+	    print row[0], row[1]
+
+if __name__ == "__main__":
+	main()

hw/hw1/starter-code/wrapper.py

+import xml.sax
+import re
+
+
+class DBLPContentHandler(xml.sax.ContentHandler):
+  """
+  Reads the dblp.xml file and produces two output files.
+        pubFile.txt = (key, pubtype) tuples
+        fieldFile.txt = (key, fieldCnt, field, value) tuples
+  Each file is tab-separated
+
+  Once the program finishes,  load these two files in a relational database; run createSchema.sql
+  """
+
+  def __init__(self):
+    xml.sax.ContentHandler.__init__(self)
+
+
+  def startElement(self, name, attrs):
+    if name == "dblp":
+      DBLPContentHandler.pubFile = open('pubFile.txt', 'w')
+      DBLPContentHandler.fieldFile = open('fieldFile.txt', 'w')
+      DBLPContentHandler.pubList = ["article", "inproceedings", "proceedings", "book", "incollection", "phdthesis", "mastersthesis", "www"]
+      DBLPContentHandler.fieldList = ["author", "editor", "title", "booktitle", "pages", "year", "address", "journal", "volume", "number", "month", "url", "ee", "cdrom", "cite", "publisher", "note", "crossref", "isbn", "series", "school", "chapter"]
+      DBLPContentHandler.content = ""
+    if name in DBLPContentHandler.pubList:
+      DBLPContentHandler.key = attrs.getValue("key")
+      DBLPContentHandler.pub = name
+      DBLPContentHandler.fieldCount = 0
+      DBLPContentHandler.content = ""
+    if name in DBLPContentHandler.fieldList:
+      DBLPContentHandler.field = name
+      DBLPContentHandler.content = ""
+ 
+  def endElement(self, name):
+    if name in DBLPContentHandler.fieldList:
+      DBLPContentHandler.fieldFile.write(DBLPContentHandler.key)
+      DBLPContentHandler.fieldFile.write("\t")
+      DBLPContentHandler.fieldFile.write(str(DBLPContentHandler.fieldCount))
+      DBLPContentHandler.fieldFile.write( "\t")
+      DBLPContentHandler.fieldFile.write(DBLPContentHandler.field)
+      DBLPContentHandler.fieldFile.write("\t")
+      DBLPContentHandler.fieldFile.write(DBLPContentHandler.content)
+      DBLPContentHandler.fieldFile.write("\n")
+      DBLPContentHandler.fieldCount += 1
+    if name in DBLPContentHandler.pubList:
+      DBLPContentHandler.pubFile.write(DBLPContentHandler.key)
+      DBLPContentHandler.pubFile.write("\t")
+      DBLPContentHandler.pubFile.write(DBLPContentHandler.pub)
+      DBLPContentHandler.pubFile.write("\n")
+
+  def characters(self, content):
+    DBLPContentHandler.content += content.replace('\\','\\\\')
+      
+def main(sourceFileName):
+  source = open(sourceFileName)
+  xml.sax.parse(source, DBLPContentHandler())
+ 
+if __name__ == "__main__":
+  main("dblp.xml")

hw/hw1/submission/README

+put your .sql files in this directory, one file per question.