Table of Contents

• 1. search
  • 1.1. Syntax
  • 1.2. Stack Effects
  • 1.3. Description
  • 1.4. PostScript Level
  • 1.5. Examples
  • 1.6. Common Use Cases
    • 1.6.1. String Parsing
    • 1.6.2. Text Replacement
    • 1.6.3. Pattern Detection
  • 1.7. Common Pitfalls
  • 1.8. Error Conditions
  • 1.9. Implementation Notes
  • 1.10. Performance Considerations
  • 1.11. Advanced Example
  • 1.12. See Also

1. search

Searches for the first occurrence of a substring within a string.

1.1. Syntax

string seek search → post match pre true (if found)
                   → string false (if not found)

1.2. Stack Effects

Table 1. Before

Level	Object
1	string (string to search in)
0	seek (substring to find)

Table 2. After (if match found)

Level	Object
3	true
2	pre (portion before match)
1	match (matching substring)
0	post (portion after match)

Table 3. After (if not found)

Level	Object
1	false
0	string (original, unchanged)

1.3. Description

search looks for the first occurrence of seek within string.

If found, it splits string into three segments: * pre: portion before the match * match: portion matching seek * post: remainder after the match

All three substrings share intervals of the original string value.

If not found, it returns the original string and false.

1.4. PostScript Level

Level 1 and later

1.5. Examples

Finding substring at start

(abbc) (ab) search
% Result: (bc) (ab) () true
% pre is empty because match at start

Finding substring in middle

(abbc) (bb) search
% Result: (c) (bb) (a) true

Finding substring at end

(abbc) (bc) search
% Result: () (bc) (ab) true
% post is empty because match at end

No match - case sensitive

(abbc) (B) search
% Result: (abbc) false

1.6. Common Use Cases

1.6.1. String Parsing

line (:) search {
  exch pop exch pop  % Keep only post
} {
  pop  % No delimiter found
} ifelse

1.6.2. Text Replacement

text (old) search {
  % Found at position
  pre (new) post
  % Concatenate: pre + new + post
} {
  pop % text unchanged
} ifelse

1.6.3. Pattern Detection

input (ERROR) search {
  pop pop pop
  true  % Error detected
} {
  pop
  false  % No error
} ifelse

1.7. Common Pitfalls

Case Sensitive - The search is exact, including case.

(Hello) (hello) search  % false - case mismatch

Finds Only First Match - Stops at the first occurrence, not all occurrences.

(abcabc) (ab) search  % Finds first 'ab' only

Stack Order - Results are post, match, pre (reversed from read order).

(hello) (ll) search  % true pre match post
% Actually: true (he) (ll) (o)

Use anchorsearch for Prefixes - If you only need to check the start of a string, anchorsearch is simpler.

1.8. Error Conditions

Error	Condition
[invalidaccess]	String has no-access attribute
[stackoverflow]	Not enough room on stack for results
[stackunderflow]	Fewer than 2 operands on stack
[typecheck]	Either operand is not a string

1.9. Implementation Notes

• Returns substrings that share the original string value

• Search is performed byte-by-byte

• Efficient for small search patterns

• Empty seek always matches at position 0

1.10. Performance Considerations

• Linear search - O(n*m) worst case

• Fast for typical use cases

• For repeated searches, consider preprocessing

• Finding all occurrences requires loops

1.11. Advanced Example

Finding all occurrences

/findall {  % string pattern => array-of-positions
  [
  3 1 roll  % [ string pattern
  0         % [ string pattern position
  {
    2 copy search {
      % [ ... string pattern pos post match pre true
      pop pop                % [ ... string pattern pos post
      3 index length         % [ ... string pattern pos post len
      4 index length sub     % [ ... string pattern pos post (string.len - post.len)
      4 -1 roll pop          % [ ... string pattern (string.len - post.len) pos
      add                    % [ ... string pattern new-pos
      dup                    % [ ... string pattern pos pos
      4 -1 roll exch         % [ ... pos string pattern pos
    } {
      pop exit               % [ ... string pattern pos
    } ifelse
  } loop
  pop pop pop                % [ positions...
  ]
} def

(the cat and the dog) (the) findall  % [0 13]

1.12. See Also

• anchorsearch - Test for prefix match

• token - Parse token from string

• getinterval - Extract substring