remove %miksearch, anchor %mikmatch at end by default

- restore `%pcre` to default anchoring

Author: zazedd

MIK.md

@@ -175,32 +175,7 @@ function%mikmatch
 This match expression will compile all of the REs in the branches into one, and use marks to find which branch was executed.  
 Efficient if you have multiple branches.
 
-#### `match%miksearch` and `function%miksearch` (search, not anchored)
-
-The previous extension was **anchored**, meaning, it will only match at the beginning of the string.
-
-This version is not, meaning, for example:
-
-```ocaml
-let mik_test = function%mikmatch
-  | {|/ (digit+ as num) /|} -> ...
-  ...
-  | _ -> failwith "no match"
-
-let () = mik_test "123" ...     (* match *)
-let () = mik_test "test123" ... (* ERROR: no match *)
-
-(* but, with %miksearch... *)
-let miks_test = function%miksearch
-  | {|/ (digit+ as num) /|} -> ...
-  ...
-  | _ -> failwith "no match"
-
-let () = miks_test "123" ...     (* match *)
-let () = miks_test "test123" ... (* match *)
-```
-
-Similar for `%miksearch_i`, except it is case insensitive.
+The regexes are anchored both at the beginning, and at the end. So, for example, the first match case will be compiled to `^some regex$`.
 
 #### General match/function
 
@@ -210,13 +185,8 @@ function
   | {%mikmatch|/ some regex /|} -> ...
   ...
   | "another string" -> ...
-  | {%miksearch|/ some regex /|} -> ... (* non-anchored *)
-  ...
-  | "yet another string" -> ...
-  | {%mikmatch_i|/ another regex /|} -> ... (* case insensitive *)
+  | {%mikmatch_i|/ some regex /|} -> ...
   ...
-  | "would you guess it" -> ...
-  | {%miksearch_i|/ another regex /|} -> ... (* non-anchored, case insensitive *)
   | _ -> ...
 ```
 

README.md

@@ -60,7 +60,7 @@ string.
 
 ### `%mikmatch`
 
-Full [%mikmatch guide](./MIK.md)
+Full [%mikmatch guide](./MIK.md).
 
 #### Quick Links
 - [Variable capture](./MIK.md#variable-capture)

common/mik_parser.mly

@@ -58,7 +58,11 @@ let unclosed_error what startpos endpos =
 %%
 
 main_match_case:
-  | SLASH p = pattern SLASH EOF { p }
+  | SLASH p = pattern SLASH EOF {
+      let dollar = to_pcre_regex "$" $endpos(p) $endpos($3) in
+      let loc = make_loc $startpos(p) $endpos($3) in
+      simplify_seq ~loc [p; dollar]
+    }
   | SLASH pattern EOF { unclosed_error "pattern (missing closing '/')" $startpos($1) $endpos }
   | SLASH error { syntax_error "Invalid pattern after opening slash" $startpos($2) $endpos($2) }
   | error { syntax_error "Expected pattern to start with '/'" $startpos($1) $endpos($1) }
@@ -137,7 +141,7 @@ basic_atom:
       to_pcre_regex $1 $startpos $endpos
     }
   | EMPTY_STR {
-      to_pcre_regex "^$" $startpos $endpos
+      to_pcre_regex "" $startpos $endpos
     }
   | UNDERSCORE {
       to_pcre_regex "." $startpos $endpos

ppx_regexp/ppx_regexp.ml

@@ -21,7 +21,6 @@ let transformation ctx =
   object (self)
     inherit [value_binding list] Ast_traverse.fold_map as super
 
-    (* Replace the entire method! structure_item in ast_builder.ml with this: *)
     method! structure_item item acc =
       match item.pstr_desc with
       (* let%mik/%pcre x = {|some regex|}*)
@@ -67,32 +66,21 @@ let transformation ctx =
         | _ -> Util.error ~loc "[%%pcre] and [%%mik] only apply to match, function and global let declarations of strings."
       in
       match e_ext.pexp_desc with
-      (* match%mik/match%pcre and function%mik/function%pcre, anchored *)
+      (* match%mikmatch/match%pcre and function%mikmatch/function%pcre, mikmatch anchored *)
       | Pexp_extension ({ txt = ("pcre" | "mikmatch" | "pcre_i" | "mikmatch_i") as ext; _ }, PStr [ { pstr_desc = Pstr_eval (e, _); _ } ])
         ->
-        let mode = if String.starts_with ~prefix:"pcre" ext then `Pcre else `Mik in
-        let opts =
-          if String.ends_with ~suffix:"_i" ext then `Caseless :: `Anchored :: Util.default_opts else `Anchored :: Util.default_opts
-        in
-        let loc = e.pexp_loc in
-        make_transformations ~mode ~opts ~loc e.pexp_desc
-      (* match%miks/match%pcres and function%miks/function%pcres, non anchored (search) *)
-      | Pexp_extension
-          ({ txt = ("pcres" | "miksearch" | "pcres_i" | "miksearch_i") as ext; _ }, PStr [ { pstr_desc = Pstr_eval (e, _); _ } ]) ->
-        let mode = if String.starts_with ~prefix:"pcre" ext then `Pcre else `Mik in
-        let opts = if String.ends_with ~suffix:"_i" ext then `Caseless :: Util.default_opts else Util.default_opts in
+        let mode, opts = if String.starts_with ~prefix:"pcre" ext then `Pcre, [] else `Mik, Util.mikmatch_default_opts in
+        let opts = if String.ends_with ~suffix:"_i" ext then `Caseless :: opts else opts in
         let loc = e.pexp_loc in
         make_transformations ~mode ~opts ~loc e.pexp_desc
-      (* match smth with | {%mik|some regex|} -> ...*)
+      (* match smth with | {%mikmatch|some regex|} -> ...*)
       | Pexp_match (matched_expr, cases) ->
         let has_ext_case =
           List.exists
             begin
               fun case ->
                 match case.pc_lhs.ppat_desc with
-                | Ppat_extension
-                    ({ txt = "pcre" | "pcres" | "mikmatch" | "miksearch" | "pcre_i" | "pcres_i" | "mikmatch_i" | "miksearch_i"; _ }, _) ->
-                  true
+                | Ppat_extension ({ txt = "pcre" | "mikmatch" | "pcre_i" | "mikmatch_i"; _ }, _) -> true
                 | _ -> false
             end
             cases
@@ -104,9 +92,7 @@ let transformation ctx =
             begin
               fun case ->
                 match case.pc_lhs.ppat_desc with
-                | Ppat_extension
-                    ({ txt = "pcre" | "pcres" | "mikmatch" | "miksearch" | "pcre_i" | "pcres_i" | "mikmatch_i" | "miksearch_i"; _ }, _) ->
-                  true
+                | Ppat_extension ({ txt = "pcre" | "mikmatch" | "pcre_i" | "pcres_i" | "mikmatch_i"; _ }, _) -> true
                 | _ -> false
             end
             cases

ppx_regexp/transformations.ml

@@ -133,7 +133,6 @@ let rec create_opts ~loc = function
   | [] -> [%expr []]
   | `Caseless :: xs -> [%expr `Caseless :: [%e create_opts ~loc xs]]
   | `Anchored :: xs -> [%expr `Anchored :: [%e create_opts ~loc xs]]
-  | `Dollar_endonly :: xs -> [%expr `Dollar_endonly :: [%e create_opts ~loc xs]]
 
 let extract_bindings ~(parser : ?pos:position -> string -> string Regexp_types.t) ~ctx ~pos s =
   let r = parser ~pos s in
@@ -328,29 +327,17 @@ let transform_mixed_match ~loc ~ctx ?matched_expr cases acc =
           (* anchored *)
           PStr [ { pstr_desc = Pstr_eval ({ pexp_desc = Pexp_constant (Pconst_string (pat, str_loc, _)); _ }, _); _ } ] ) ->
       let pos = str_loc.loc_start in
-      let mode = if String.starts_with ~prefix:"pcre" ext then `Pcre else `Mik in
-      let opts =
-        if String.ends_with ~suffix:"_i" ext then `Caseless :: `Anchored :: Util.default_opts else `Anchored :: Util.default_opts
-      in
-      let parser = match mode with `Pcre -> Regexp.parse_exn ~target:`Match | `Mik -> Regexp.parse_mik_exn ~target:`Match in
-      let re, bs, nG = extract_bindings ~parser ~pos ~ctx pat in
-      `Mik (opts, re, nG, bs, case.pc_rhs, case.pc_guard)
-    | Ppat_extension
-        ( { txt = ("pcres" | "miksearch" | "pcres_i" | "miksearch_i") as ext; _ },
-          (* search, non anchored *)
-          PStr [ { pstr_desc = Pstr_eval ({ pexp_desc = Pexp_constant (Pconst_string (pat, str_loc, _)); _ }, _); _ } ] ) ->
-      let pos = str_loc.loc_start in
-      let mode = if String.starts_with ~prefix:"pcre" ext then `Pcre else `Mik in
-      let opts = if String.ends_with ~suffix:"_i" ext then `Caseless :: Util.default_opts else Util.default_opts in
+      let mode, opts = if String.starts_with ~prefix:"pcre" ext then `Pcre, [] else `Mik, Util.mikmatch_default_opts in
+      let opts = if String.ends_with ~suffix:"_i" ext then `Caseless :: opts else opts in
       let parser = match mode with `Pcre -> Regexp.parse_exn ~target:`Match | `Mik -> Regexp.parse_mik_exn ~target:`Match in
       let re, bs, nG = extract_bindings ~parser ~pos ~ctx pat in
-      `Mik (opts, re, nG, bs, case.pc_rhs, case.pc_guard)
+      `Ext (opts, re, nG, bs, case.pc_rhs, case.pc_guard)
     | _ -> `Regular case
   in
 
   let prepared_cases = List.map aux cases in
 
-  let has_mik = List.exists (function `Mik _ -> true | _ -> false) prepared_cases in
+  let has_mik = List.exists (function `Ext _ -> true | _ -> false) prepared_cases in
 
   if not has_mik then begin
     match matched_expr with None -> pexp_function ~loc cases, acc | Some m -> pexp_match ~loc m cases, acc
@@ -361,7 +348,7 @@ let transform_mixed_match ~loc ~ctx ?matched_expr cases acc =
         begin
           fun i case ->
             match case with
-            | `Mik (opts, re, _, _, _, _) ->
+            | `Ext (opts, re, _, _, _, _) ->
               let comp_var = Util.fresh_var () in
               let opts_expr = create_opts ~loc opts in
               let comp_expr = [%expr Re.compile (Re.Perl.re ~opts:[%e opts_expr] [%e re])] in
@@ -385,7 +372,7 @@ let transform_mixed_match ~loc ~ctx ?matched_expr cases acc =
           match [%e input_var] with
           | [%p case.pc_lhs] when [%e Option.value case.pc_guard ~default:[%expr true]] -> [%e case.pc_rhs]
           | _ -> [%e build_ordered_match input_var (case_idx + 1) rest mik_comps]]
-      | `Mik (_, _, _, bs, rhs, guard) :: rest, (idx, comp_var, _) :: rest_comps when idx = case_idx ->
+      | `Ext (_, _, _, bs, rhs, guard) :: rest, (idx, comp_var, _) :: rest_comps when idx = case_idx ->
         let comp_ident = pexp_ident ~loc { txt = Lident comp_var; loc } in
         [%expr
           match Re.exec_opt [%e comp_ident] [%e input_var] with
@@ -398,7 +385,7 @@ let transform_mixed_match ~loc ~ctx ?matched_expr cases acc =
                 let guarded_rhs = [%expr if [%e g] then [%e rhs] else [%e build_ordered_match input_var (case_idx + 1) rest rest_comps]] in
                 wrap_group_bindings ~captured_acc:[] ~loc guarded_rhs 0 bs]
           | None -> [%e build_ordered_match input_var (case_idx + 1) rest rest_comps]]
-      | `Mik _ :: rest, _ ->
+      | `Ext _ :: rest, _ ->
         (* shouldn't happen if indices are correct *)
         build_ordered_match input_var (case_idx + 1) rest mik_comps
     in

ppx_regexp/util.ml

@@ -1,7 +1,7 @@
 open Ppxlib
 open Ast_builder.Default
 
-let default_opts = [ `Dollar_endonly ]
+let mikmatch_default_opts = [ `Anchored ]
 let error = Location.raise_errorf
 
 let warn ~loc msg e =