HTTP Selected Headers

Extract selected headers from a raw HTTP request.

request_lines holds the request split into lines. Line 0 is always the request line (for example "GET /v1/charges HTTP/1.1") and is skipped. Header lines follow until the first empty string ""; everything after it (the body) is ignored. wanted lists header names to extract.

Each remaining line is one of:

• Continuation line: starts with a space or tab. It extends the most recent header line. Never treat it as a new header, even if it contains :.
• Header line: contains :. The name is the text before the first :, trimmed; the value is the text after it.
• Malformed line: anything else, or a header line whose trimmed name is empty. It is ignored, and continuation lines right after it are ignored too.

Rules:

• Header names match wanted case-insensitively; output names are lowercase.
• Normalize every value by trimming it and collapsing runs of spaces/tabs to a single space.
• A continuation line appends its normalized text to the previous header's value, separated by one space (no leading space if the value was empty). Whitespace-only continuations add nothing. Continuations of headers that are not wanted are dropped.
• Repeated headers are reported separately.

Return [lowercase_name, value] pairs for the wanted headers in the order they appear.