Refresh demo to Rebiber 1.3.0

Install Rebiber from GitHub main, pin Gradio 5.49.1, copy the latest
abbr.tsv, and fix Gradio 4/5 button updates. Keep matching local-only
(no live DBLP lookup) with an optional skip-author-check toggle.

README.md

@@ -4,9 +4,15 @@ emoji: 📚
 colorFrom: blue
 colorTo: blue
 sdk: gradio
-sdk_version: 5.29.1
+sdk_version: 5.49.1
 app_file: app.py
 pinned: true
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Rebiber 1.3.0 demo
+
+Normalize BibTeX with official DBLP / ACL Anthology records.
+
+This Space installs [Rebiber](https://github.com/yuchenlin/rebiber) from GitHub `main` (latest dumps) and runs **local-only** matching — no live DBLP lookup.
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

abbr.tsv

@@ -3,9 +3,10 @@ Proc. of NAACL-HLT | Proceedings of the .* Conference of the North {A}merican Ch
 Proc. of EMNLP | Proceedings of the .* Conference on Empirical Methods in Natural Language Processing
 Proc. of ACL | Proceedings of the .* Annual Meeting of the Association for Computational Linguistics
 Proc. of WWW | {WWW} '.*: The Web Conference .*
-Proc. of ICLR | .* International Conference on Learning Representations,
-Proc. of ICML | Proceedings of the .* International Conference on Machine Learning.*
+Proc. of ICLR | .*International Conference on Learning Representations
+Proc. of ICML | .*International Conference on Machine Learning
 Proc. of ICML | Machine Learning, Proceedings of the .* International Conference
+Proc. of NeurIPS | .*Neural Information Processing Systems
 Proc. of AAAI | Proceedings of the .*AAAI.* Conference on Artificial Intelligence
 Proc. of KDD | The .* {ACM} {SIGKDD} International Conference on Knowledge Discovery
 Proc. of CVPR | .*IEEE Computer Society Conference on Computer Vision and Pattern Recognition.*
@@ -13,4 +14,23 @@ Proc. of CVPR | .*IEEE/CVF Conference on Computer Vision and Pattern Recognition
 Proc. of ICDE | .*International Conference on Data Engineering.*
 Proc. of SIGIR | .*International {ACM} {SIGIR}.*
 Proc. of INTERSPEECH | Interspeech .*Annual Conference of the International Speech.*
-TKDE | IEEE Transactions on Knowledge and Data Engineering
+TKDE | IEEE Transactions on Knowledge and Data Engineering
+Proc. of COLM | .*Conference on Language Modeling
+Proc. of WACV | .*Winter Conference on Applications of Computer Vision
+Proc. of ECCV | .*European Conference on Computer Vision
+Proc. of ICCV | .*International Conference on Computer Vision
+Proc. of NAACL | Proceedings of the .* Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics
+Proc. of Findings | Findings of the Association for Computational Linguistics
+TACL | Transactions of the Association for Computational Linguistics
+JMLR | Journal of Machine Learning Research
+TMLR | Trans. Mach. Learn. Res.
+TMLR | Transactions on Machine Learning Research
+JMLR | J. Mach. Learn. Res.
+Proc. of MICCAI | .*Medical Image Computing and Computer.?Assisted Intervention
+Proc. of MLSys | .*Conference on Machine Learning and Systems
+Proc. of LoG | .*Learning on Graphs Conference
+Proc. of AutoML | .*International Conference on Automated Machine Learning
+Proc. of ICRA | .*International Conference on Robotics and Automation
+Proc. of IROS | .*International Conference on Intelligent Robots and Systems
+Proc. of RSS | .*Robotics: Science and Systems
+Proc. of CoRL | .*Conference on Robot Learning

app.py

@@ -1,32 +1,44 @@
 import os
+import tempfile
 import uuid
+
 import gradio as gr
 import rebiber
+from rebiber.normalize import load_abbr_tsv
+
+# ---------- preload DB (local dumps only; no live DBLP) ---------- #
+PACKAGE_ROOT = os.path.dirname(os.path.abspath(rebiber.__file__))
+SPACE_ROOT = os.path.dirname(os.path.abspath(__file__))
 
-# ---------- preload DB ---------- #
-PACKAGE_ROOT = os.path.abspath(rebiber.__file__).replace("__init__.py", "")
 bib_db = rebiber.construct_bib_db(
     os.path.join(PACKAGE_ROOT, "bib_list.txt"), start_dir=PACKAGE_ROOT
 )
-abbr_dict = rebiber.normalize.load_abbr_tsv("abbr.tsv")
+
+abbr_path = os.path.join(SPACE_ROOT, "abbr.tsv")
+if not os.path.isfile(abbr_path):
+    abbr_path = os.path.join(PACKAGE_ROOT, "abbr.tsv")
+abbr_dict = load_abbr_tsv(abbr_path)
+
 
 # ---------- helper functions ---------- #
-def process(input_bib: str,
-            shorten: bool,
-            remove_keys: list[str],
-            deduplicate: bool,
-            sort_ids: bool):
+def process(
+    input_bib: str,
+    shorten: bool,
+    remove_keys,
+    deduplicate: bool,
+    sort_ids: bool,
+    skip_author_check: bool,
+):
     """Normalize a BibTeX string and write it to disk."""
-    if "@" not in input_bib:
-        # Nothing that looks like BibTeX
-        return "N/A", "", gr.Button.update(visible=False)
+    if not input_bib or "@" not in input_bib:
+        return "N/A", "", gr.update(visible=False), gr.update(visible=False, value=None)
 
     run_id = uuid.uuid4().hex
-    in_file  = f"input_{run_id}.bib"
-    out_file = f"output_{run_id}.bib"
+    tmpdir = tempfile.gettempdir()
+    in_file = os.path.join(tmpdir, f"rebiber_input_{run_id}.bib")
+    out_file = os.path.join(tmpdir, f"rebiber_output_{run_id}.bib")
 
-    # Write raw input
-    with open(in_file, "w") as f:
+    with open(in_file, "w", encoding="utf8") as f:
         f.write(input_bib.replace("\t", "    "))
 
     entries = rebiber.load_bib_file(in_file)
@@ -34,24 +46,35 @@ def process(input_bib: str,
         bib_db,
         entries,
         out_file,
-        abbr_dict = abbr_dict if shorten else [],
-        deduplicate = deduplicate,
-        sort = sort_ids,
-        removed_value_names = remove_keys,
+        abbr_dict=abbr_dict if shorten else [],
+        deduplicate=bool(deduplicate),
+        sort=bool(sort_ids),
+        removed_value_names=list(remove_keys or []),
+        check_authors=not bool(skip_author_check),
+        live_lookup=False,
     )
 
-    with open(out_file) as f:
+    with open(out_file, encoding="utf8") as f:
         output_bib = f.read().replace("\n ", "\n    ")
 
-    # Show output + enable “Download” button
-    return output_bib, run_id, gr.update(visible=True)
+    return output_bib, out_file, gr.update(visible=True), gr.update(visible=False, value=None)
 
 
-def download_file(run_id: str):
+def download_file(out_path: str):
     """Expose the normalized .bib for download."""
-    file_path = f"output_{run_id}.bib"
-
-    return file_path, gr.update(visible=True)
+    if not out_path or not os.path.isfile(out_path):
+        return gr.update(visible=False, value=None)
+    return gr.update(value=out_path, visible=True)
+
+
+def clear_all():
+    return (
+        "",
+        "",
+        "",
+        gr.update(visible=False),
+        gr.update(visible=False, value=None),
+    )
 
 
 # ---------- demo UI ---------- #
@@ -67,19 +90,21 @@ EXAMPLE = """
 with gr.Blocks(title="Rebiber", theme=gr.themes.Soft()) as demo:
     gr.Markdown(
         """
-# **Rebiber** – normalise those BibTeX entries!  
+# **Rebiber 1.3.0** – normalise those BibTeX entries!
 
 🐼 [Project page](https://yuchenlin.xyz/)   |  
 :octocat: [GitHub](https://github.com/yuchenlin/rebiber)   |  
-🐤 [Tweet](https://twitter.com/billyuchenlin/status/1353850378438070272)  
+🐤 [Tweet](https://twitter.com/billyuchenlin/status/1353850378438070272)
 
-Rebiber replaces arXiv citations with their official venue versions (DBLP / ACL
-Anthology), deduplicates, sorts, and can abbreviate conference names.
+Rebiber **1.3.0** replaces arXiv citations with official venue versions (DBLP / ACL Anthology),
+deduplicates, sorts, and can abbreviate conference names. Matching is **local-only** (no live DBLP lookup).
+
+**What's new in 1.3.0:** author-overlap guard (a shared title like "Deep Learning" is not replaced by an unrelated paper),
+official arXiv fields (`eprint`, `archivePrefix`, `primaryClass`), and the latest conference dumps.
 """
     )
 
     with gr.Row():
-        # ---------- left column ---------- #
         with gr.Column(scale=3):
             input_bib = gr.Textbox(
                 label="Input BibTeX",
@@ -88,41 +113,61 @@ Anthology), deduplicates, sorts, and can abbreviate conference names.
                 interactive=True,
             )
             removekeys = gr.CheckboxGroup(
-                ["url", "biburl", "address", "publisher",
-                 "pages", "doi", "volume", "bibsource"],
+                [
+                    "url",
+                    "biburl",
+                    "address",
+                    "publisher",
+                    "pages",
+                    "doi",
+                    "volume",
+                    "bibsource",
+                ],
                 label="Remove fields",
-                info="Select the keys you’d like to strip",
+                info="Select the keys you'd like to strip",
+            )
+            shorten = gr.Checkbox(label="Abbreviate venue names", value=False)
+            dedup = gr.Checkbox(label="Deduplicate entries", value=False)
+            sort = gr.Checkbox(label="Sort IDs alphabetically", value=False)
+            skip_author_check = gr.Checkbox(
+                label="Skip author check",
+                value=False,
+                info="Disable the 1.3 author-overlap guard (check_authors=False). Leave unchecked unless you know you need it.",
             )
-            shorten = gr.Checkbox(label="Abbreviate venue names")
-            dedup   = gr.Checkbox(label="Deduplicate entries")
-            sort    = gr.Checkbox(label="Sort IDs alphabetically")
 
             with gr.Row():
-                clear_btn   = gr.Button("Clear")
-                submit_btn  = gr.Button("Submit")
+                clear_btn = gr.Button("Clear")
+                submit_btn = gr.Button("Submit", variant="primary")
 
-            run_uuid = gr.Textbox(visible=False)
+            run_path = gr.Textbox(visible=False)
 
-        # ---------- right column ---------- #
         with gr.Column(scale=3):
             output_box = gr.Textbox(
                 label="Normalised BibTeX",
                 interactive=False,
-                show_copy_button=True,      # :contentReference[oaicite:2]{index=2}
+                show_copy_button=True,
+                lines=15,
             )
-            download_btn = gr.Button("Generate . bib file", visible=False)
-            download_file_component = gr.File(visible=False)
+            download_btn = gr.Button("Download .bib", visible=False)
+            download_file_component = gr.File(label="Normalized BibTeX file", visible=False)
 
-    # ---------- wiring ---------- #
     submit_btn.click(
         process,
-        inputs=[input_bib, shorten, removekeys, dedup, sort],
-        outputs=[output_box, run_uuid, download_btn],
+        inputs=[input_bib, shorten, removekeys, dedup, sort, skip_author_check],
+        outputs=[output_box, run_path, download_btn, download_file_component],
         api_name="process",
     )
-    download_btn.click(download_file, run_uuid, [download_file_component, download_file_component])
-    clear_btn.click(lambda: "", None, input_bib)
+    download_btn.click(
+        download_file,
+        inputs=run_path,
+        outputs=download_file_component,
+    )
+    clear_btn.click(
+        clear_all,
+        inputs=None,
+        outputs=[input_bib, output_box, run_path, download_btn, download_file_component],
+    )
 
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()

requirements.txt

@@ -1 +1,2 @@
-git+https://github.com/yuchenlin/rebiber.git@main
+git+https://github.com/yuchenlin/rebiber.git@main
+gradio==5.49.1