new build file

build_bib.py

@@ -0,0 +1,213 @@
+import sys
+import os
+import argparse
+import pandas as pd
+import logging
+
+# constants
+DELIMITER = "$"
+NEW_LINE = "\n"
+SAVE_SUFFIX = "_generated_py3.bib"
+
+# key tracker
+REF_KEY_DICT = {}
+
+# filenames
+VENUE = "venue_fullname.txt"
+CONFERENCE = "conference.txt"
+JOURNAL = "journal.txt"
+ARXIV = "arxiv.txt"
+BOOK = "book.txt"
+WEBSITE = "website.txt"
+
+
+def build_ref_key(bib_item):
+
+    author_list = bib_item["author"].strip().split(" and ")
+    ref_key = ""
+    for a in author_list:
+        name_part = a.strip().split(" ")
+        ref_key += name_part[-1][0]
+    ref_key += str(bib_item["year"])[2:]
+
+    return ref_key
+
+
+def process_title(df):
+    df["title"] = df["title"].str.replace("<dollar-inline>", "$")
+
+
+def postprocess_refkey(ref_key):
+    if ref_key in REF_KEY_DICT:
+        REF_KEY_DICT[ref_key] += 1
+        ref_key += str(REF_KEY_DICT[ref_key])
+    else:
+        REF_KEY_DICT[ref_key] = 1
+    return ref_key
+
+
+def identify_normal_proceedings(basename, venue):
+    match venue:
+        case "WWW_old":
+            proceedings = basename + " (WWW)"
+        case "KDD_old":
+            proceedings = basename + " (KDD)"
+        case "ICWSM_old":
+            proceedings = basename + " (ICWSM)"
+        case "NeurIPSCT":
+            proceedings = basename
+        case _:
+            proceedings = basename + f" ({venue})"
+    return proceedings
+
+
+def build_conference(f, mode):
+
+    venue_fullname = pd.read_csv(VENUE, sep=DELIMITER)
+    conference = pd.read_csv(CONFERENCE, sep=DELIMITER)
+    process_title(conference)
+
+    for i in range(conference.shape[0]):
+
+        ref_key = build_ref_key(conference.loc[i])
+
+        if ref_key:
+            ref_key = postprocess_refkey(ref_key)
+
+            venue = conference.loc[i, "venue"]
+            if mode == "normal":
+                basename = venue_fullname.loc[
+                    venue_fullname.venue == venue, "fullname"
+                ].values[0]
+                proceedings = identify_normal_proceedings(basename, venue)
+            elif mode == "simple":
+                if venue == "WWW_old":
+                    proceedings = "WWW"
+                else:
+                    proceedings = venue
+
+            bib_item = f"""@inproceedings{{{ref_key},{NEW_LINE}"""
+            bib_item += f"""author = {{{conference.loc[i, "author"]}}},{NEW_LINE}"""
+            bib_item += f"""title = {{{conference.loc[i, "title"]}}},{NEW_LINE}"""
+            bib_item += f"""booktitle = {{{proceedings}}},{NEW_LINE}"""
+
+            if not isinstance(conference.loc[i, "pages"], float):
+                bib_item += f"""pages = {{{conference.loc[i, "pages"]}}},{NEW_LINE}"""
+
+            if mode == "normal":
+                if not isinstance(conference.loc[i, "publisher"], float):
+                    bib_item += f"""publisher = {{{conference.loc[i, "publisher"]}}},{NEW_LINE}"""
+            elif mode == "simple":
+                bib_item += f"""publisher = {{}},{NEW_LINE}"""
+
+            bib_item += f"""year = {{{conference.loc[i, "year"]}}}{NEW_LINE}"""
+            bib_item += f"""}}{NEW_LINE}"""
+
+            f.write(bib_item)
+            f.write("\n")
+
+
+def build_journal(f, mode):
+
+    venue_fullname = pd.read_csv(VENUE, sep=DELIMITER)
+    journal = pd.read_csv(JOURNAL, sep=DELIMITER)
+    process_title(journal)
+
+    for i in range(journal.shape[0]):
+
+        ref_key = build_ref_key(journal.loc[i])
+
+        if ref_key:
+            ref_key = postprocess_refkey(ref_key)
+
+            proceedings = venue_fullname.loc[
+                venue_fullname.venue == journal.loc[i, "venue"], "fullname"
+            ].values[0]
+
+            bib_item = f"""@article{{{ref_key},{NEW_LINE}"""
+            bib_item += f"""author = {{{journal.loc[i, "author"]}}},{NEW_LINE}"""
+            bib_item += f"""title = {{{journal.loc[i, "title"]}}},{NEW_LINE}"""
+            bib_item += f"""journal = {{{proceedings}}},{NEW_LINE}"""
+            if not isinstance(journal.loc[i, "publisher"], float):
+                bib_item += (
+                    f"""publisher = {{{journal.loc[i, "publisher"]}}},{NEW_LINE}"""
+                )
+            bib_item += f"""year = {{{journal.loc[i, "year"]}}}{NEW_LINE}"""
+            bib_item += f"""}}{NEW_LINE}"""
+            f.write(bib_item)
+            f.write("\n")
+
+
+def build_arxiv(f, mode):
+    arxiv = pd.read_csv(ARXIV, sep=DELIMITER)
+    process_title(arxiv)
+
+    for i in range(arxiv.shape[0]):
+
+        ref_key = build_ref_key(arxiv.loc[i])
+
+        if ref_key:
+            ref_key = postprocess_refkey(ref_key)
+
+            bib_item = f"""@article{{{ref_key},{NEW_LINE}"""
+            bib_item += f"""author = {{{arxiv.loc[i, "author"]}}},{NEW_LINE}"""
+            bib_item += f"""title = {{{arxiv.loc[i, "title"]}}},{NEW_LINE}"""
+            bib_item += f"""journal = {{{{{arxiv.loc[i, "venue"] }}}}},{NEW_LINE}"""
+            bib_item += f"""year = {{{arxiv.loc[i, "year"]}}}{NEW_LINE}"""
+            bib_item += f"""}}{NEW_LINE}"""
+            f.write(bib_item)
+            f.write("\n")
+
+
+def build_book(f, mode):
+    book = pd.read_csv(BOOK, sep=DELIMITER)
+    process_title(book)
+
+    for i in range(book.shape[0]):
+
+        ref_key = build_ref_key(book.loc[i])
+
+        if ref_key:
+            ref_key = postprocess_refkey(ref_key)
+
+            bib_item = f"""@book{{{ref_key},{NEW_LINE}"""
+            bib_item += f"""author = {{{book.loc[i, "author"]}}},{NEW_LINE}"""
+            bib_item += f"""title = {{{book.loc[i, "title"]}}},{NEW_LINE}"""
+            bib_item += f"""publisher = {{{book.loc[i, "publisher"]}}},{NEW_LINE}"""
+            bib_item += f"""year = {{{book.loc[i, "year"]}}}{NEW_LINE}"""
+            bib_item += f"""}}{NEW_LINE}"""
+            f.write(bib_item)
+            f.write("\n")
+
+
+def build_website(f, mode):
+    website = pd.read_csv(WEBSITE, sep=DELIMITER)
+    for i in range(website.shape[0]):
+
+        bib_item = f"""@misc{{{website.loc[i, "key"]},{NEW_LINE}"""
+        bib_item += f"""howpublished = {{\\url{{{website.loc[i, "url"]}}}}}{NEW_LINE}"""
+        bib_item += f"""}}{NEW_LINE}"""
+        f.write(bib_item)
+        f.write("\n")
+
+
+def build(mode="normal"):
+    with open(mode + SAVE_SUFFIX, "a") as f:
+        build_conference(f, mode)
+        build_journal(f, mode)
+        build_arxiv(f, mode)
+        build_book(f, mode)
+        build_website(f, mode)
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--mode", default="normal", help="bibtex mode (simple/normal)")
+
+    args = parser.parse_args()
+    mode = args.mode
+    if os.path.exists(mode + SAVE_SUFFIX):
+        os.remove(mode + SAVE_SUFFIX)
+
+    build(mode)