feat: file info dataframe from filenames and file content

[MthwRobinson]

Summary

Adds functions for building a dataframe of file info from a list of filenames or a list of file contents. Supports creating this plot for our snazzy demo UI.

Testing

Run the following Python code.

import os
from unstructured.file_utils.exploration import get_file_info

filenames = [os.path.join("example-docs/", f) for f in os.listdir("example-docs/")]
get_file_info(filenames)

You should get a dataframe that looks like this:

                         filename  ...       filetype
0                  fake-html.html  ...  FileType.HTML
1                example-10k.html  ...  FileType.HTML
2                    factbook.xml  ...   FileType.XML
3           fake-email-header.eml  ...   FileType.UNK
4                       fake.docx  ...  FileType.DOCX
5   fake-email-image-embedded.eml  ...   FileType.EML
6                   fake-text.txt  ...   FileType.TXT
7    layout-parser-paper-fast.pdf  ...   FileType.PDF
8            email-with-image.eml  ...   FileType.EML
9           fake-power-point.pptx  ...  FileType.PPTX
10                 fake-email.txt  ...   FileType.TXT
11                      README.md  ...   FileType.TXT
12                   factbook.xsl  ...   FileType.XML
13                fake-excel.xlsx  ...  FileType.XLSX
14                 fake-email.eml  ...   FileType.EML
15        layout-parser-paper.pdf  ...   FileType.PDF
16      fake-email-attachment.eml  ...   FileType.EML
17                    example.jpg  ...   FileType.JPG

[18 rows x 5 columns]

[cragwolfe]

tested locally on a dir that included empty file, works for me!

Though, the following is emitted, which isn't super helpful as it doesn't say what file is not supported.

MIME type was inode/x-empty. This file type is not currently supported in unstructured.

[MthwRobinson]

Thanks, added #208 to capture updating that warning to make it more helpful.