Capture Algorithm

Reqflow parses documents sequentially and works line by line.

Reqflow considers the following as a single line:

• a line of a raw text or PDF document
• a <p> paragraph of an XML based document (HTML, docx, odt)

Reqflow expects the document to be structured as a sequence of requirements with mixed text, where a requirement REQ is structured as follows:

REQ        := LINE_REQ
              [ LINES_TEXT ]
              [ REFS ]

LINE_REQ   := REQ_ID [ REFS ]

REQ_ID     := the unique identifier of the requirement

LINES_TEXT := some text, on one or more lines

REFS       := REF [ REF ... ]

REF        := the unique identifier of a reference

As a result:

• a REQ_ID is always before its REF.
• a REF is always associated to the REQ_ID before it

The Reqflow parameters are as follows:

• -req tells how to capture the REQ. Parentheses may be used to identify REQ inside a broader expression.
  • Eg: <(REQ_[-a-zA-Z_0-9]*)> for matching <REQ_123> and extracting REQ_123
• -ref tells how to capture the REFS. Parentheses may be used to identify one or more REFS inside a broader expression.
  • Eg: Ref:[, ]*(REF_[0-9]+) for matching Ref: REF_01, REF_02, REF_03 and extracting REF_01, REF_02, REF_03
  • Eg: REF_[0-9]+ for matching REF_01, and extracting the same.
• -end-req tells where the capture of the text of the requirement shall end