需要帮助来构建正则表达式

Question

I am trying to find a regex for the following lines (see attempts at bottom of post)

CLog_DMT_HPCC2_IWHT91731695_242_AFT1_2019-05-02T07.51.43

The Regex is working fine for this line, Result for above line is:

• programName=CLog
• otherRegex=DMT_HPCC2
• SerialNO=IWHT91731695 (Note: Serial no will always start from "I")
• Version=242
• operation=AFT1

which is desired result

But the Regex is not working for this line

CLOB_ABCD_6KW_SYSTEM_609-784_IWHT91831863_197_ACB_01_2019-05-02T07.03.27

I want result in following way for above line

• programName=CLOB
• otherRegex=ABCD_6KW_SYSTEM_609-784
• SerialNO=IWHT91831863
• Version= 197
• operation= ACB_01

but what i am getting is shown below:

• programName=CLOB
• otherRegex=ABCD_6KW_SYSTEM_609-784
• SerialNO=IWHT91831863 _197
• Version= ACB
• operation= 01

I have tried following regex for above lines:

(?<programName>[a-zA-Z0-9]+)_(?<other>.+)_(?<boardSN>I.+)_(?<entityNameProgramVersion>.+)_(?<operation>.+)_

Answer 1

In your pattern you use .+ which is greedy and will match until the end of the string. Then it will backtrack to fulfill the rest of the pattern. In this case it will try backtrack to fit all the following underscores.

Instead you could use a negated character class [^ matching not an underscore or a newline to limit the matching to the current string in case when there are multiple following.

For the other part you might make the quantifier non greedy (?<other>.+?) so it gives up matches until it can match _I

^(?<programName>[a-zA-Z0-9]+)_(?<other>.+?)_(?<boardSN>I[^_\n]+)_(?<entityNameProgramVersion>[^_\n]+)_(?<operation>[^\n_]+(?:_[^\n]+)?)_

Explanation

• ^ Start of string
• (?<programName>[a-zA-Z0-9]+)_ Repeat 1+ times what is listed in the character class
• (?<other>.+?)_ Match any char 1+ times except a newline non greedy
• (?<boardSN>I[^_\\n]+)_ Negated character class, match not an _ or newline
• (?<entityNameProgramVersion>[^_\\n]+)_ Negated character class, match not an _ or newline
• (?<operation>[^\\n_]+(?:_[^\\n]+)?)_ Negated character class, match not an _ or newline with an optional group which will match one underscore and match not an underscore. After that, match a single underscore outside of the group.

Regex demo

If optional group at the end can be only digits, you could use this part without the last underscore:

(?<operation>[^\n_]+(?:_\d+)?)

Regex demo