File Type Breakdown vs unzip -l and 7z l

The classic awk one-liner vs the CSV

The shell pipeline counts by extension but is case-sensitive and includes directory lines. File Type Breakdown lower-cases and excludes folders, so the grouped output is cleaner.

Shell:
  $ unzip -l assets.zip | awk -F. 'NF>1{print tolower($NF)}' | sort | uniq -c
     12 png
      3 PNG     <-- without tolower these split
      8 js

File Type Breakdown (assets-types.csv):
  extension,count,uncompressedSize,compressedSize,ratio
  png,15,...   <-- .PNG and .png already merged
  js,8,...
  (folders excluded automatically)

Getting compressed sizes: zipinfo vs the ratio column

To get per-file compressed sizes in the shell you reach for zipinfo or unzip -v and parse columns. File Type Breakdown gives a per-extension compressed total and ratio directly.

Shell (per-file, must aggregate yourself):
  $ zipinfo -l build.zip
  ... (defl:N) columns you then sum per extension ...

File Type Breakdown (build-types.csv):
  extension,count,uncompressedSize,compressedSize,ratio
  js,318,9842110,2110874,78.6%
  png,142,24180736,23905111,1.1%
  No column-parsing — the ratio is computed for you.

7z and RAR without installing p7zip / unrar

On a managed laptop without p7zip you cannot run 7z l. File Type Breakdown reads 7z and RAR via libarchive WASM in the browser, no install needed — though sizes are post-decompression so the ratio reads 0.0%.

Shell (needs p7zip installed):
  $ 7z l release.7z

File Type Breakdown (release-types.csv):
  extension,count,uncompressedSize,compressedSize,ratio
  dll,84,40221884,40221884,0.0%
  exe,3,9920114,9920114,0.0%
  ratio is 0.0% because 7z is decompressed before counting.

Why Total entries differs from unzip -l's line count

unzip -l prints a row per entry including directories and a summary line. File Type Breakdown's Total entries counts files only. Expect the tool's number to be lower for archives with many folders.

Shell:
  $ unzip -l deep-tree.zip | tail -1
     1240 files
  (but the listing also showed 86 directory rows)

File Type Breakdown metrics:
  Distinct types: 9  |  Total entries: 1240
  Directories (86) are excluded from the type rows entirely.

Scripting a batch the tool cannot do

For a nightly job profiling 200 archives, the shell wins — loop and aggregate. The tool is one-archive, manual. Use the CLI here and reserve the tool for ad-hoc inspection.

Shell (batch — out of scope for the browser tool):
  for f in /artifacts/*.zip; do
    echo "== $f"
    unzip -l "$f" | awk -F. 'NF>1{print tolower($NF)}' \
      | sort | uniq -c | sort -rn
  done

File Type Breakdown: drop one archive at a time; no loop/API.

A naive unzip -l | wc -l includes directory rows and the summary line. File Type Breakdown excludes directories, so its Total entries is files-only and will usually be smaller. Neither is wrong — they count different things.

Without tr A-Z a-z or tolower(), the shell pipeline reports .JPG and .jpg as separate types. File Type Breakdown always lower-cases, so those merge. Add lower-casing to your pipeline to match.

Archives above the tier ceiling (2 GB on the highest tiers, 50 MB free) are rejected by the tool. The CLI streams and has no such cap, so for very large archives the shell is the right choice.

TAR has no per-file compressed figure — neither tar -tvf nor File Type Breakdown can show one. For a TAR.GZ the tool's compressed column equals the uncompressed column, matching what the CLI can know.

The CLI path for 7z and RAR requires installing p7zip and unrar. File Type Breakdown reads both via libarchive WASM with nothing to install — handy on locked-down machines.

There is no API or multi-archive mode for this tool — it is one archive per run. For batch profiling across many files, script the CLI instead.

Both approaches can be fooled by extension alone, but File Type Breakdown sniffs magic bytes, so a RAR named .zip is read correctly. Confirm with auto-format-detector if unsure.

The tool's CSV is aggregated per extension; zipinfo is per file. The aggregate is intentional — to match zipinfo line-for-line you would sum its rows by extension yourself.