Open the appIndex Extractor

Once you finish tagging a manuscript, the next problem arrives.
“So how many personal names are in this book, exactly?”
Have you ever counted tags by eye to build that list?
Honestly, half of index work is making this list.

This tool pulls out only what sits inside your chosen symbols and turns it into a list.
Place names wrapped in 【 】, index terms wrapped in <index>, words emphasized with **.
Even if the same thing appears a hundred times it stays once in the list, and you can see how many times it appeared.

How to Use

Basic Steps

  1. Paste your manuscript into the Source Text box.
  2. Write the opening and closing symbols in Extraction Symbols.
  3. Click [Extract Index Terms].

Writing the Symbols

Separate the opening and closing symbol with a single tab character.

opening[tab]closing

Write several lines and results are grouped by symbol.

【	】
<index>	</index>
**	**

Identical Opening and Closing Symbols Work Too

Even when both sides are the same, like '' or **, the pairs are found correctly.

''	''
**	**

From a manuscript like this

贈鄭寅協''庚申(1920)''
惟聖性者·**佛氏**心者

you get this.

# ''	''
庚申(1920)

# **	**
佛氏

Adding Comments

You can leave notes in the symbol settings too. Use # or //.

<index>	</index>  #tagged index
【	】  // place names

Here’s the interesting part — the comments show up in the result as well.

# <index>	</index>  #tagged index
dog
cat

# 【	】  // place names
Gyeongju
Seoul

Extraction strips the comment and uses only the symbols, but the result header carries your note along.
Even when the list grows long you know which block is which.
The :// in http:// is not treated as a comment, so URLs are safe.

Choosing the Sort Order

Turn on Show counts and occurrences are counted, sorted most frequent first.

# 【	】
48	Bruce Lee
18	Lee Kyung-kyu
5	Kang Ho-dong

Turn it off and you get plain alphabetical order.

# 【	】
Kang Ho-dong
Lee Kyung-kyu
Bruce Lee

Alphabetical is convenient when building an index; counts are useful when judging who matters.
Someone appearing 48 times carries a different weight than someone appearing 5 times.

Formatting a Translation Institute Collation List

A dedicated mode for handling collation material from the Institute for the Translation of Korean Classics.
Check it and the tool ignores your symbol input and uses < and > fixed. It works even with the symbol box empty.

The heart of this mode is cutting off everything after ^.

<庚申^collation note>  →  庚申

Collation material is written as term^note, but the list only needs the front part.
Duplicates are cleaned again after cutting, so even if one term carries several notes it appears once.

Sorting differs too.
Instead of scattering things alphabetically or by count, it keeps the order they appear in the original.
With collation material, the order itself is information.

Good to Know

The Numbers in the Result

Three figures appear at the bottom after extraction.

ItemMeaning
Extracted termsHow many remain after removing duplicates
Duplicates removedHow many overlapped and disappeared
Symbol pairsHow many pairs were actually used

If you wrote three pairs but “symbol pairs” says 1, the other two never matched anything in the manuscript.
Either the symbols are wrong or the manuscript doesn’t use them — worth checking.

Hanja Is Matched for You

Multi-code Hanja normalization is applied automatically to both the source text and the symbols.
The same character failing to match because of a different code simply doesn’t happen.

Pairs Well with the Index Tagger

Tag your manuscript with the Index Tagger, then pull those tags back out with this tool and your index list is done.
Tagging and extraction are two halves of one job.

Upload

Click the Upload button in each app to load text files.

Supported formats: .txt, .md, .html, .xml, .tsv, .csv

Save

Download the converted result with the Save button.

Copy

Copy the result to your clipboard with the Copy button.

Open the appIndex Extractor