thinking regularly, thinking universally, thinking mathematically

List of Patterns 🎇 'Pattern-Collector'

The 'awesome List' of Patterns *(the only repo of it's kind)
Please edit this draft wildy 🎉: Spreadsheet / Readme ) Please don't hesitate to add sublists for specific scientific fields such as DNA
Exploration: Patterns make bite-sized tools🍒🍟 ( Searching such list (,once well populated,) you will already have mentally defined the specific, regular scope of your goal (the task of identifying specific data / matches). That can be more efficient / versatile than searching Stack Overflow Answers or node.js NPM's. Yet each regex could also be an NPM or module/package in any language.

---

1. Regular Expressions(=Search Patterns=Data format definitions.)

Regex are most common & most efficient to type. (Despite they are one of the oldest dicsiplines in programming to make sense of data, convert it, clean it or spell-check it. https://en.wikipedia.org/wiki/Regular_expression)
Regex are versatile, because they work in most languages and editors and many apps.

Common Data Formats²	match	replacement	comment/justify	extra³_
ISBN
Youtube Video ID	[^\w-]([\w-]{11})[^\w-]	$1	11char base64 is almost unique	(?:https?://|//)?(?:www\.|m\.)?youtu/?be(?:\.com)?/(?:embed/|v/|watch\/?\?[&\w=]{,128}v=([\w-]{11})[^\w-]
Hashes, Public Keys, Signatures	match
MD6
SHA256, Bitcoin, ...
Convert	match	replacement
MarkDown links to HTML links	\[([^\]]*)\]\(([^\)]*)\)	<a href="\$2">\$1</a>
this table2Javascript	\	`([^`]*)`\|`([^`]*)`\		`replaceAll(/$1/g, "$2").replaceAll("\
Javascript 2 Python	...	$12\3

² date, postal code, formal greeting, formal __, ...
³extra: match typos too (common) and/or add precision ('no false positives' / perfectionism)

[we could add 1000s]

---

1.1 Automatic pattern generation / AI

Currently (2021) little of this is automated. Solutions such as Microsoft Power Automate for Desktop (Windows 11) want to change some of it.

1.2 Pre-processing Patterns

A raw text / data source material - or a list or category of patterns - can sometimes be analyzed for similarities and thus be combined in one preprocessing step. i.e. Preprocessing might Reduce Input data by 90% already in a fraction of the time / CPU

2. Contextual & Semantic patterns

word-lists, topics, frequencies, thesaurus, antonyms, semantic dictionaries, psychologic & sentiment dictionaries

wordnet, framenet, google ngrams, google trends, ....

Google Search:

~synonyms a|b AROUND(3) c|d -e|f|g|h|i|j|k|l|m|n|o|p|q|r|s|t|u|v|w|x|y|z
https://ahrefs.com/blog/google-advanced-search-operators/

Human Grammar & Natural language processing (NLP):

https://github.com/edobashira/speech-language-processing#readme

3. Structured Data. Querying Public Databases & the internet. SPARQL, SQL, NoSQL

Semantic web

WikiData

AWS public databases

4. Merging the above "1.-3."

vs 5. Human work VS machine learning models

---

All Patterns

https://docs.google.com/spreadsheets/d/1EjeZ2RtNpM_mANdO1VPXmZmbIb5vANUXodPBFtdg3zU/edit

• Others Lists // potential Sources: ___ , ___ , ___ , ____ ,____ , ( not a list but 1 repo per regex: https://github.com/regexhq, takes clicks to see one: regexhq/youtube-regex/index.js)

• Compare: https://www.mulesoft.com/exchange/?type=connector&view=list (>10000 'enterprise converts')

Name	pattern match	replacement	language	comment/justify	raw³	extra context/precision
			regex
			google
			css