Scripts for generating ISMIR 2018 proceedings
Directory | Purpose |
---|---|
./ |
Scripts for global processing |
data/ |
Input: Metadata about papers, sessions, etc |
2018_Proceedings_ISMIR/ |
LaTeX files for generating PDF proceedings |
2018_Proceedings_ISMIR/articles/ |
Input: PDF files of all articles by number, e.g., 004_Paper.pdf |
2018_Proceedings_ISMIR/external/ |
Input: Front matter created in Word or google docs |
2018_Proceedings_ISMIR_Electronic_Tools/ |
Scripts for generating HTML proceedings for USB sticks |
2018_Proceedings_ISMIR_Electronic_Tools/templates/ |
Input: templates for HTML proceedings, DBLP |
2018_Proceedings_ISMIR_Electronic/ |
Output: some HTML proceedings files |
2018_Proceedings_ISMIR_Electronic_Tools/output/ |
Output: other HTML proceedings files |
- Install NotePad++, Python 3.6, Perl, Tex editor
- Import paper metadata (paper list, reviewers, program committee, session info, and session index) into
data/completePaperList.xlsx
- Check the consistency of paper titles and author names between the meta data and the camera-ready PDF
- Check the consistency of author names of the same author
- Order paper titles in each poster session in alphabetical order
- Order reviewers by their last name in alphabetical order
- Order program committee members by their last name in alphabetical order
- sessionInfo sheet: column A is the session tag; column B is the session name displayed in the proceedings; column C is the page number of the '2018_Proceedings_ISMIR/external/12_Sessions.docx'
- session_index sheet: Start Page column can be specified later once the proceedings is assembled. It's used for extracting the stamped PDF of each paper for the electronic proceedings.
- Save each of the above sheets as a .txt file with the same name as the sheet in 'data/' folder using utf8 encoding; To do so in Windows, first save the sheet as a unicode text file, then use Notepad++ to change its encoding.
- Run
./deriveFiles.py
to derive other metadata files from those. They include:- One .txt file for each session in the '2018_Proceedings_ISMIR_Electronic_Tools\data' folder
- 2018_Proceedings_ISMIR/papers.tex
- Create directory
2018_Proceedings_ISMIR/articles/
and put all camera-ready PDFs inside - Run
2018_Proceedings_ISMIR/01-get_pages_total.py
to calculate page lengths of files for indexing purposes. This file will be saved to '2018_Proceedings_ISMIR_Electronic_Tools\data' folder - Modify '2018_Proceedings_ISMIR/confproc.cls to change the header to the new conference
- Modify files in the '2018_Proceedings_ISMIR/external/' folder accordingly
- Start cmd as administrator; Run
2018_Proceedings_ISMIR/00-run.sh
to compile the PDF proceedings. It creates three copies, in '2018_Proceedings_ISMIR', '2018_Proceedings_ISMIR_Electronic', and '2018_Proceedings_ISMIR_Electronic_Tools\data', respectively.- On Windows, I encountered some problems running 00-run.sh. I had to install GitBash to run the "sed" command, and use cmd to run "pdflatex" and "authorindex" commands;
- To run the "authorindex" command, you need to download Perl.
- You may need to manually correct a couple of author names in the generated author index list, as the splitting of first/last names by "authorindex" can be wrong.
- Modify session_index sheet of 'data/completePaperList.xlsx' to reflect the true starting page numbers of sessions, then save this sheet as 'session_index.txt' to the 'data/' folder with utf8 edcoding. Copy this file to '2018_Proceedings_ISMIR_Electronic_Tools/data'
- Run
2018_Proceedings_ISMIR_Electronic_Tools/00-split_proceedings.py
. Each paper's PDF with the ISMIR proceedings stamp is now splitted into '2018_Proceedings_ISMIR_Electronic_Tools\data\articles_splitted' - Modify files '2018_Proceedings_ISMIR_Electronic_Tools/templates', especially the dblp.txt file
- Run
2018_Proceedings_ISMIR_Electronic_Tools/01-generate_overview_website.py
- Run
2018_Proceedings_ISMIR_Electronic_Tools/02-generate_author_index.py
- Run
2018_Proceedings_ISMIR_Electronic_Tools/03-generate_reviewer_list.py
- Compose a website for USB key content based on the generated pages; the cover page requires some photos of the attractions in Suzhou.