Function frequency
You are encouraged to solve this task according to the task description, using any language you may know.
Display - for a program or runtime environment (whatever suits the style of your language) - the top ten most frequently occurring functions (or also identifiers or tokens, if preferred).
This is a static analysis: The question is not how often each function is actually executed at runtime, but how often it is used by the programmer.
Besides its practical usefulness, the intent of this task is to show how to do self-inspection within the language.
ACL2
This will, unfortunately, also catch variable names after an open paren, such as in (let ...) expressions.
<lang Lisp>(in-package "ACL2")
(set-state-ok t)
(defun read-all-objects (limit channel state)
(mv-let (eof obj state)
(read-object channel state)
(if (or eof (zp limit))
(mv nil state)
(mv-let (so-far state)
(read-all-objects (- limit 1) channel state)
(mv (cons obj so-far) state)))))
(defun list-starters (xs)
(cond ((endp xs) nil)
((consp (first xs))
(append (if (symbolp (first (first xs)))
(list (first (first xs)))
nil)
(list-starters (rest (first xs)))
(list-starters (rest xs))))
(t (list-starters (rest xs)))))
(defun invoked-functions (filename state)
(mv-let (channel state)
(open-input-channel filename :object state)
(mv-let (code state)
(read-all-objects 1000 channel state)
(mv (list-starters code) state))))
(defun increment-for (key alist)
(cond ((endp alist) (list (cons key 1)))
((equal key (car (first alist)))
(cons (cons key (1+ (cdr (first alist))))
(rest alist)))
(t (cons (first alist)
(increment-for key (rest alist))))))
(defun symbol-freq-table (symbols)
(if (endp symbols)
nil
(increment-for (first symbols)
(symbol-freq-table (rest symbols)))))
(defun insert-freq-table (pair alist)
(cond ((endp alist)
(list pair))
((> (cdr pair) (cdr (first alist)))
(cons pair alist))
(t (cons (first alist)
(insert-freq-table pair (rest alist))))))
(defun isort-freq-table (alist)
(if (endp alist)
nil
(insert-freq-table (first alist)
(isort-freq-table (rest alist)))))
(defun main (state)
(mv-let (fns state)
(invoked-functions "function-freq.lisp" state)
(mv (take 10 (isort-freq-table
(symbol-freq-table fns))) state)))</lang>
Output (for itself):
(((FIRST . 10) (REST . 8) (DEFUN . 8) (CONS . 7) (MV-LET . 5) (LIST-STARTERS . 4) (IF . 4) (MV . 4) (COND . 3) (LIST . 3)) <state>)
AWK
<lang AWK>
- syntax: GAWK -f FUNCTION_FREQUENCY.AWK filename(s).AWK
- sorting:
- PROCINFO["sorted_in"] is used by GAWK
- SORTTYPE is used by Thompson Automation's TAWK
BEGIN {
- create array of keywords to be ignored by lexer
asplit("BEGIN:END:atan2:break:close:continue:cos:delete:" \
"do:else:exit:exp:for:getline:gsub:if:in:index:int:" \
"length:log:match:next:print:printf:rand:return:sin:" \
"split:sprintf:sqrt:srand:strftime:sub:substr:system:tolower:toupper:while",
keywords,":")
- build the symbol-state table
split("00:00:00:00:00:00:00:00:00:00:" \
"20:10:10:12:12:11:07:00:00:00:" \
"08:08:08:08:08:33:08:00:00:00:" \
"08:44:08:36:08:08:08:00:00:00:" \
"08:44:45:42:42:41:08",machine,":")
- parse the input
state = 1
for (;;) {
symb = lex() # get next symbol
nextstate = substr(machine[state symb],1,1)
act = substr(machine[state symb],2,1)
# perform required action
if (act == "0") { # do nothing
}
else if (act == "1") { # found a function call
if (!(inarray(tok,names))) {
names[++nnames] = tok
}
++xnames[tok]
}
else if (act == "2") { # found a variable or array
if (tok in Local) {
tok = tok "(" funcname ")"
if (!(inarray(tok,names))) {
names[++nnames] = tok
}
++xnames[tok]
}
else {
tok = tok "()"
if (!(inarray(tok,names))) {
names[++nnames] = tok
}
++xnames[tok]
}
}
else if (act == "3") { # found a function definition
funcname = tok
}
else if (act == "4") { # found a left brace
braces++
}
else if (act == "5") { # found a right brace
braces--
if (braces == 0) {
delete Local
funcname = ""
nextstate = 1
}
}
else if (act == "6") { # found a local variable declaration
Local[tok] = 1
}
else if (act == "7") { # found end of file
break
}
else if (act == "8") { # found an error
printf("error: FILENAME=%s, FNR=%d\n",FILENAME,FNR)
exit(1)
}
state = nextstate # finished with current token
}
- format function names
for (i=1; i<=nnames; i++) {
if (index(names[i],"(") == 0) {
tmp_arr[xnames[names[i]]][names[i]] = ""
}
}
- print function names
PROCINFO["sorted_in"] = "@ind_num_desc" ; SORTTYPE = 9
for (i in tmp_arr) {
PROCINFO["sorted_in"] = "@ind_str_asc" ; SORTTYPE = 1
for (j in tmp_arr[i]) {
if (++shown <= 10) {
printf("%d %s\n",i,j)
}
}
}
exit(0)
} function asplit(str,arr,fs, i,n,temp_asplit) {
n = split(str,temp_asplit,fs)
for (i=1; i<=n; i++) {
arr[temp_asplit[i]]++
}
} function inarray(val,arr, j) {
for (j in arr) {
if (arr[j] == val) {
return(j)
}
}
return("")
} function lex() {
for (;;) {
if (tok == "(eof)") {
return(7)
}
while (length(line) == 0) {
if (getline line == 0) {
tok = "(eof)"
return(7)
}
}
sub(/^[ \t]+/,"",line) # remove white space,
sub(/^"([^"]|\\")*"/,"",line) # quoted strings,
sub(/^\/([^\/]|\\\/)+\//,"",line) # regular expressions,
sub(/^#.*/,"",line) # and comments
if (line ~ /^function /) {
tok = "function"
line = substr(line,10)
return(1)
}
else if (line ~ /^{/) {
tok = "{"
line = substr(line,2)
return(2)
}
else if (line ~ /^}/) {
tok = "}"
line = substr(line,2)
return(3)
}
else if (match(line,/^[A-Za-z_][A-Za-z_0-9]*\[/)) {
tok = substr(line,1,RLENGTH-1)
line = substr(line,RLENGTH+1)
return(5)
}
else if (match(line,/^[A-Za-z_][A-Za-z_0-9]*\(/)) {
tok = substr(line,1,RLENGTH-1)
line = substr(line,RLENGTH+1)
if (!(tok in keywords)) { return(6) }
}
else if (match(line,/^[A-Za-z_][A-Za-z_0-9]*/)) {
tok = substr(line,1,RLENGTH)
line = substr(line,RLENGTH+1)
if (!(tok in keywords)) { return(4) }
}
else {
match(line,/^[^A-Za-z_{}]/)
tok = substr(line,1,RLENGTH)
line = substr(line,RLENGTH+1)
}
}
} </lang>
Output of running FUNCTION_FREQUENCY.AWK on itself:
3 inarray 1 asplit 1 lex
Sample input:
BEGIN {
f1()
f2();f2()
f3a();f3a();f3a()
f3b();f3b();f3b()
f4();f4();f4();f4()
f5();f5();f5();f5();f5()
exit(0)
}
function f0() { }
function f1() { }
function f2() { }
function f3a() { }
function f3b() { }
function f4() { }
function f5() { }
Sample output:
5 f5 4 f4 3 f3a 3 f3b 2 f2 1 f1
BBC BASIC
<lang bbcbasic> INSTALL @lib$+"SORTLIB"
Sort% = FN_sortinit(1,0) : REM Descending
Valid$ = "0123456789@ABCDEFGHIJKLMNOPQRSTUVWXYZ_`abcdefghijklmnopqrstuvwxyz"
DIM func$(1000), cnt%(1000)
nFunc% = 0
file% = OPENIN("*.bbc")
WHILE NOT EOF#file%
ll% = BGET#file%
no% = BGET#file% + 256*BGET#file%
INPUT #file%, l$
i% = 1
REPEAT
j% = INSTR(l$, CHR$&A4, i%) : REM Token for 'FN'
k% = INSTR(l$, CHR$&F2, i%) : REM Token for 'PROC'
IF k% IF j%=0 OR j%>k% THEN
i% = k%
f$ = "PROC"
ELSE
i% = j%
f$ = "FN"
ENDIF
IF i% THEN
REPEAT
i% += 1
f$ += MID$(l$, i%, 1)
UNTIL INSTR(Valid$, MID$(l$, i%+1, 1)) = 0
FOR j% = 0 TO nFunc%-1
IF f$ = func$(j%) EXIT FOR
NEXT
IF j% >= nFunc% nFunc% += 1
func$(j%) = f$
cnt%(j%) += 1
ENDIF
UNTIL i%=0
ENDWHILE
CLOSE #file%
C% = nFunc%
CALL Sort%, cnt%(0), func$(0)
IF C% > 10 C% = 10
FOR i% = 0 TO C%-1
PRINT func$(i%) " (" ; cnt%(i%) ")"
NEXT</lang>
Output (for file LBB.BBC):
FNcheck (291) FNexpr (125) FNtoken (49) PROCout (41) FNhandle (31) FNupper (30) FNitem (30) FNcheckns (21) FNinstrq (17) FNchild (17)
C
This program treats doesn't differentiate between macros and functions. It works by looking for function calls which are not inside strings or comments. If a function call has a C style comment between the opening brace and the name of the function, this program will not recognize it as a function call. <lang C>
- define _POSIX_SOURCE
- include <ctype.h>
- include <stdio.h>
- include <stdlib.h>
- include <errno.h>
- include <string.h>
- include <stddef.h>
- include <sys/mman.h>
- include <sys/types.h>
- include <sys/stat.h>
- include <unistd.h>
struct functionInfo {
char* name; int timesCalled; char marked;
}; void addToList(struct functionInfo** list, struct functionInfo toAdd, \
size_t* numElements, size_t* allocatedSize)
{
static const char* keywords[32] = {"auto", "break", "case", "char", "const", \
"continue", "default", "do", "double", \
"else", "enum", "extern", "float", "for", \
"goto", "if", "int", "long", "register", \
"return", "short", "signed", "sizeof", \
"static", "struct", "switch", "typedef", \
"union", "unsigned", "void", "volatile", \
"while"
};
int i;
/* If the "function" being called is actually a keyword, then ignore it */
for (i = 0; i < 32; i++) {
if (!strcmp(toAdd.name, keywords[i])) {
return;
}
}
if (!*list) {
*allocatedSize = 10;
*list = calloc(*allocatedSize, sizeof(struct functionInfo));
if (!*list) {
printf("Failed to allocate %lu elements of %lu bytes each.\n", \
*allocatedSize, sizeof(struct functionInfo));
abort();
}
(*list)[0].name = malloc(strlen(toAdd.name)+1);
if (!(*list)[0].name) {
printf("Failed to allocate %lu bytes.\n", strlen(toAdd.name)+1);
abort();
}
strcpy((*list)[0].name, toAdd.name);
(*list)[0].timesCalled = 1;
(*list)[0].marked = 0;
*numElements = 1;
} else {
char found = 0;
unsigned int i;
for (i = 0; i < *numElements; i++) {
if (!strcmp((*list)[i].name, toAdd.name)) {
found = 1;
(*list)[i].timesCalled++;
break;
}
}
if (!found) {
struct functionInfo* newList = calloc((*allocatedSize)+10, \
sizeof(struct functionInfo));
if (!newList) {
printf("Failed to allocate %lu elements of %lu bytes each.\n", \
(*allocatedSize)+10, sizeof(struct functionInfo));
abort();
}
memcpy(newList, *list, (*allocatedSize)*sizeof(struct functionInfo));
free(*list);
*allocatedSize += 10;
*list = newList;
(*list)[*numElements].name = malloc(strlen(toAdd.name)+1);
if (!(*list)[*numElements].name) {
printf("Failed to allocate %lu bytes.\n", strlen(toAdd.name)+1);
abort();
}
strcpy((*list)[*numElements].name, toAdd.name);
(*list)[*numElements].timesCalled = 1;
(*list)[*numElements].marked = 0;
(*numElements)++;
}
}
} void printList(struct functionInfo** list, size_t numElements) {
char maxSet = 0;
unsigned int i;
size_t maxIndex = 0;
for (i = 0; i<10; i++) {
maxSet = 0;
size_t j;
for (j = 0; j<numElements; j++) {
if (!maxSet || (*list)[j].timesCalled > (*list)[maxIndex].timesCalled) {
if (!(*list)[j].marked) {
maxSet = 1;
maxIndex = j;
}
}
}
(*list)[maxIndex].marked = 1;
printf("%s() called %d times.\n", (*list)[maxIndex].name, \
(*list)[maxIndex].timesCalled);
}
} void freeList(struct functionInfo** list, size_t numElements) {
size_t i;
for (i = 0; i<numElements; i++) {
free((*list)[i].name);
}
free(*list);
} char* extractFunctionName(char* readHead) {
char* identifier = readHead;
if (isalpha(*identifier) || *identifier == '_') {
while (isalnum(*identifier) || *identifier == '_') {
identifier++;
}
}
/* Search forward for spaces and then an open parenthesis
* but do not include this in the function name.
*/
char* toParen = identifier;
if (toParen == readHead) return NULL;
while (isspace(*toParen)) {
toParen++;
}
if (*toParen != '(') return NULL;
/* Copy the found function name to the output string */
ptrdiff_t size = (ptrdiff_t)((ptrdiff_t)identifier) \
- ((ptrdiff_t)readHead)+1;
char* const name = malloc(size);
if (!name) {
printf("Failed to allocate %lu bytes.\n", size);
abort();
}
name[size-1] = '\0';
memcpy(name, readHead, size-1);
/* Function names can't be blank */
if (strcmp(name, "")) {
return name;
}
free(name);
return NULL;
} int main(int argc, char** argv) {
int i;
for (i = 1; i<argc; i++) {
errno = 0;
FILE* file = fopen(argv[i], "r");
if (errno || !file) {
printf("fopen() failed with error code \"%s\"\n", \
strerror(errno));
abort();
}
char comment = 0;
- define DOUBLEQUOTE 1
- define SINGLEQUOTE 2
int string = 0;
struct functionInfo* functions = NULL;
struct functionInfo toAdd;
size_t numElements = 0;
size_t allocatedSize = 0;
struct stat metaData;
errno = 0;
if (fstat(fileno(file), &metaData) < 0) {
printf("fstat() returned error \"%s\"\n", strerror(errno));
abort();
}
char* mmappedSource = (char*)mmap(NULL, metaData.st_size, PROT_READ, \
MAP_PRIVATE, fileno(file), 0);
if (errno) {
printf("mmap() failed with error \"%s\"\n", strerror(errno));
abort();
}
if (!mmappedSource) {
printf("mmap() returned NULL.\n");
abort();
}
char* readHead = mmappedSource;
while (readHead < mmappedSource + metaData.st_size) {
while (*readHead) {
/* Ignore comments inside strings */
if (!string) {
if (*readHead == '/' && !strncmp(readHead, "/*", 2)) {
comment = 1;
}
if (*readHead == '*' && !strncmp(readHead, "*/", 2)) {
comment = 0;
}
}
/* Ignore strings inside comments */
if (!comment) {
if (*readHead == '"') {
if (!string) {
string = DOUBLEQUOTE;
} else if (string == DOUBLEQUOTE) {
/* Only toggle string mode if the quote character
* is not escaped
*/
if (strncmp((readHead-1), "\\\"", 2)) {
string = 0;
}
}
}
if (*readHead == '\) {
if (!string) {
string = SINGLEQUOTE;
} else if (string == SINGLEQUOTE) {
if (strncmp((readHead-1), "\\\'", 2)) {
string = 0;
}
}
}
}
/* Look for identifiers outside of any comment or string */
if (!comment && !string) {
char* name = extractFunctionName(readHead);
/* Don't read part of an identifier on the next iteration */
if (name) {
toAdd.name = name;
addToList(&functions, toAdd, &numElements, &allocatedSize);
readHead += strlen(name);
}
free(name);
}
readHead++;
}
}
errno = 0;
munmap(mmappedSource, metaData.st_size);
if (errno) {
printf("munmap() returned error \"%s\"\n", strerror(errno));
abort();
}
errno = 0;
fclose(file);
if (errno) {
printf("fclose() returned error \"%s\"\n", strerror(errno));
abort();
}
printList(&functions, numElements);
freeList(&functions, numElements);
}
return 0;
}</lang>
Common Lisp
Loading the file itself before scanning it is the quickest way to determine what function bindings would be created. <lang lisp>(defun mapc-tree (fn tree)
"Apply FN to all elements in TREE."
(cond ((consp tree)
(mapc-tree fn (car tree))
(mapc-tree fn (cdr tree)))
(t (funcall fn tree))))
(defun count-source (source)
"Load and count all function-bound symbols in a SOURCE file."
(load source)
(with-open-file (s source)
(let ((table (make-hash-table)))
(loop for data = (read s nil nil)
while data
do (mapc-tree
(lambda (x)
(when (and (symbolp x) (fboundp x))
(incf (gethash x table 0))))
data))
table)))
(defun hash-to-alist (table)
"Convert a hashtable to an alist." (let ((alist)) (maphash (lambda (k v) (push (cons k v) alist)) table) alist))
(defun take (n list)
"Take at most N elements from LIST." (loop repeat n for x in list collect x))
(defun top-10 (table)
"Get the top 10 from the source counts TABLE." (take 10 (sort (hash-to-alist table) '> :key 'cdr)))</lang>
- Output:
CL-USER> (top-10 (count-source "function-frequency.lisp")) ((DEFUN . 5) (MAPC-TREE . 4) (QUOTE . 2) (LIST . 2) (TAKE . 2) (HASH-TO-ALIST . 2) (LAMBDA . 2) (LOOP . 2) (LET . 2) (CDR . 2))
Erlang
This is source code analyse. Mainly because I have never done that before. <lang Erlang> -module( function_frequency ).
-export( [erlang_source/1, task/0] ).
erlang_source( File ) ->
{ok, IO} = file:open( File, [read] ),
Forms = parse_all( IO, io:parse_erl_form(IO, ), [] ),
Functions = lists:flatten( [erl_syntax_lib:fold(fun accumulate_functions/2, [], X) || X <- Forms] ),
dict:to_list( lists:foldl(fun count/2, dict:new(), Functions) ).
task() ->
Function_frequencies = erlang_source( "function_frequency.erl" ),
{Top_tens, _Rest} = lists:split( 10, lists:reverse(lists:keysort(2, Function_frequencies)) ),
[io:fwrite("Function ~p called ~p times.~n", [X, Y]) || {X, Y} <- Top_tens].
accumulate_functions( Tree, Acc ) -> accumulate_functions( erlang:element(1, Tree), Tree, Acc ).
accumulate_functions( call, Tree, Acc ) -> [accumulate_functions_name(Tree) | Acc]; accumulate_functions( _Other, _Tree, Acc ) -> Acc.
accumulate_functions_name( Tree ) -> accumulate_functions_name_scoop( erlang:element(3, Tree) ).
accumulate_functions_name_scoop( {atom, _Line, Name} ) -> Name; accumulate_functions_name_scoop( {remote, _Line, {atom, _Line, Module}, {atom, _Line, Name}} ) -> {Module, Name}.
count( Key, Dict ) -> dict:update_counter( Key, 1, Dict ).
parse_all( _IO, {eof, _End}, Acc ) -> Acc; parse_all( IO, {ok, Tokens, Location}, Acc ) -> parse_all( IO, io:parse_erl_form(IO, , Location), [Tokens | Acc] ). </lang>
- Output:
7> function_frequency:task().
Function parse_all called 2 times.
Function {erlang,element} called 2 times.
Function {io,parse_erl_form} called 2 times.
Function {lists,flatten} called 1 times.
Function {dict,update_counter} called 1 times.
Function {lists,reverse} called 1 times.
Function {lists,foldl} called 1 times.
Function {io,fwrite} called 1 times.
Function accumulate_functions_name called 1 times.
Function erlang_source called 1 times.
Factor
Let's take a look at the sequences vocabulary/source file from Factor's standard library. This approach does not count word-defining words such as : and M:, nor does it count words like { and [.
<lang factor>USING: accessors kernel math.statistics prettyprint sequences
sequences.deep source-files vocabs words ;
"resource:core/sequences/sequences.factor" "sequences" [ path>source-file top-level-form>> ] [ vocab-words [ def>> ] [ ] map-as ] bi* compose [ word? ] deep-filter sorted-histogram <reversed> 7 head .</lang>
- Output:
{
{ if 54 }
{ dup 53 }
{ drop 46 }
{ dip 44 }
{ swap 42 }
{ length 39 }
{ keep 36 }
}
Forth
Counts colon definitions, variables, constants, words defined by definers like CREATE..DOES>, etc. Check for levels of top from command line, by default 4. GForth 0.7.0 specific. <lang Forth>' noop is bootmessage
\ --- LIST OF CONSTANTS \ WORD# maximum word size \ RING# size of `Rings' element \ DEFS definitions \ KEYS \ \ --- LIST OF VARIABLES \ cmpl? is compiling? \ cword current compiled word
wordlist constant DEFS wordlist constant KEYS
\ --- Compiling 50 constant WORD#
- >>fPAD ( ca u -- ; u < 51 )
PAD 80 blank s" create " PAD swap MOVE s" 1 , DOES> 1 swap +! ;" PAD 57 + swap MOVE WORD# min PAD 7 + swap MOVE ;
- funcmpl ( ca u -- )
>>fPAD current @ DEFS current ! PAD 80 evaluate current ! ;
- >>kPAD ( ca u -- ; )
PAD 80 blank s" : " PAD swap MOVE s" parse-name funcmpl ;" PAD 59 + swap MOVE WORD# min PAD 2 + swap MOVE ;
- keycmpl ( ca u -- )
>>kPAD current @ KEYS current ! PAD 80 evaluate current ! ;
\ --- Interpreter
- intp BEGIN parse-name dup
WHILE ( ca u ) 2dup KEYS search-wordlist IF execute 2drop ELSE DEFS search-wordlist IF execute THEN THEN REPEAT 2drop ;
- run BEGIN refill WHILE intp REPEAT ;
\ --- Lists&Rings warnings OFF
- LIST ( node -- ) ]] BEGIN @ dup WHILE >R [[ ; immediate
warnings ON
- LOOP-LIST ( -- ) ]] R> REPEAT drop [[ ; immediate
- empty-ring? ( node -- f ) dup @ = ;
- RING ( node -- ) ]] dup BEGIN @ 2dup <> WHILE 2>R [[ ; immediate
- LOOP-RING ( -- ) ]] 2R> REPEAT 2drop [[ ; immediate
- new-node ( -- node )
here dup , ;
- do-link ( node new-node -- ; do link after current node )
over @ over ! swap ! ;
\ --- Sorting..
- nt>freq ( nt -- n ;frequency of uses )
name>int >BODY @ ;
- @maxfreq ( wid -- n ;maximum frequency )
0 swap cell+ LIST ( max ) I nt>freq 2dup < IF nip ELSE drop THEN LOOP-LIST ;
2 cells constant RING#
- rings-vec ( u -- a size ; create vector of rings )
here over 1+ 0 DO new-node drop 0 , LOOP swap RING# * ;
- populate-by ( a wid -- )
cell+ LIST dup I nt>freq RING# * + \ root-node new-node I , \ new-node do-link LOOP-LIST drop ;
\ --- Display TOP
- node>nt cell+ @ ;
- .ring ( root-node -- )
0 swap RING dup 0= IF I node>nt nt>freq . THEN space I node>nt name>string type 1+ LOOP-RING drop cr ;
- .top ( a size n -- )
-rot BOUNDS swap ?DO ( n ) I empty-ring? 0= IF 1- I .ring THEN dup 0= IF drop UNLOOP EXIT THEN [ RING# negate ] LITERAL +LOOP drop ;
- args>top# ( -- n )
1 arg 2dup 0 0 d<> IF >float IF f>d d>s dup 0= IF drop 4 THEN ELSE 4 THEN ELSE 2drop 4 THEN ;
\ --- KEYS behaviour
variable cmpl? cmpl? OFF 2variable cword
here WORD# allot 0 cword 2!
current @ KEYS current !
- create
cmpl? @ IF cword 2@ keycmpl ELSE parse-name funcmpl THEN ;
- constant
cmpl? @ IF cword 2@ keycmpl ELSE parse-name funcmpl THEN ;
- variable parse-name funcmpl ;
- value parse-name funcmpl ;
- defer parse-name funcmpl ;
- ( BEGIN >in @ [char] ) parse nip >in @ rot - =
WHILE refill 0= IF exit THEN REPEAT ;
- \ 10 parse 2drop ;
- \G 10 parse 2drop ;
- S" [char] " parse 2drop ;
- ." [char] " parse 2drop ;
- [']
parse-name DEFS search-wordlist IF execute THEN ;
- postpone
parse-name DEFS search-wordlist IF execute THEN ;
- ; cmpl? OFF ;
- : warnings OFF
parse-name cword 2@ drop WORD# rot umin dup >R MOVE cword 2@ drop R> cword 2! cword 2@ cmpl? @ IF keycmpl \ `:' inside def. = a defining word ELSE funcmpl THEN cmpl? ON warnings ON
current !
\ Run, ruuun! stdin ' run execute-parsing-file DEFS @maxfreq rings-vec over DEFS populate-by args>top# .top bye </lang> Self test:
$ $ gforth funfreq.fs 10 < funfreq.fs 7 DEFS funcmpl cmpl? 5 WORD# 4 KEYS keycmpl nt>freq RING# 3 LIST LOOP-LIST new-node node>nt 2 >>fPAD >>kPAD intp run empty-ring? RING LOOP-RING do-link @maxfreq rings-vec populate-by .ring .top args>top# 1 create constant variable value defer ( \ \G S" ." ['] postpone ; : $ $ gforth funfreq.fs < funfreq.fs 7 DEFS funcmpl cmpl? 5 WORD# 4 KEYS keycmpl nt>freq RING# 3 LIST LOOP-LIST new-node node>nt $
Go
Only crude approximation is currently easy in Go. The following parses source code, looks for function call syntax (an expression followed by an argument list) and prints the expression. <lang go>package main
import (
"fmt" "go/ast" "go/parser" "go/token" "io/ioutil" "os" "sort"
)
func main() {
if len(os.Args) != 2 {
fmt.Println("usage ff <go source filename>")
return
}
src, err := ioutil.ReadFile(os.Args[1])
if err != nil {
fmt.Println(err)
return
}
fs := token.NewFileSet()
a, err := parser.ParseFile(fs, os.Args[1], src, 0)
if err != nil {
fmt.Println(err)
return
}
f := fs.File(a.Pos())
m := make(map[string]int)
ast.Inspect(a, func(n ast.Node) bool {
if ce, ok := n.(*ast.CallExpr); ok {
start := f.Offset(ce.Pos())
end := f.Offset(ce.Lparen)
m[string(src[start:end])]++
}
return true
})
cs := make(calls, 0, len(m))
for k, v := range m {
cs = append(cs, &call{k, v})
}
sort.Sort(cs)
for i, c := range cs {
fmt.Printf("%-20s %4d\n", c.expr, c.count)
if i == 9 {
break
}
}
}
type call struct {
expr string count int
} type calls []*call
func (c calls) Len() int { return len(c) } func (c calls) Swap(i, j int) { c[i], c[j] = c[j], c[i] } func (c calls) Less(i, j int) bool { return c[i].count > c[j].count }</lang> Output, when run on source code above:
len 3 fmt.Println 3 f.Offset 2 make 2 fmt.Printf 1 ioutil.ReadFile 1 a.Pos 1 string 1 token.NewFileSet 1 append 1
J
The lowest approach taken here makes no effort to classify the primitives as monads nor dyads, nor as verbs, adverbs, nor conjunctions. Did "-" mean "additive inverse" or indicate subtraction? Does ";" raze or link? J is a multi-instruction single data language. Parentheses around a group of verbs form hooks or forks which affect data flow. The simple top10 verb does not find these important constructs. So we shall ignore them for this exercise.
<lang j>
IGNORE=: ;:'y(0)1',CR
Filter=: (#~`)(`:6)
NB. extract tokens from a large body newline terminated of text
roughparse=: ;@(<@;: ::("_);._2)
NB. count frequencies and get the top x
top=: top=: {. \:~@:((#;{.)/.~)
NB. read all installed script (.ijs) files and concatenate them
JSOURCE=: ;fread each 1&e.@('.ijs'&E.)@>Filter {."1 dirtree jpath '~install'
10 top (roughparse JSOURCE)-.IGNORE
┌─────┬──┐ │49591│, │ ├─────┼──┤ │40473│=:│ ├─────┼──┤ │35593│; │ ├─────┼──┤ │34096│=.│ ├─────┼──┤ │24757│+ │ ├─────┼──┤ │18726│" │ ├─────┼──┤ │18564│< │ ├─────┼──┤ │18446│/ │ ├─────┼──┤ │16984│> │ ├─────┼──┤ │14655│@ │ └─────┴──┘ </lang>
Julia
<lang julia>using Printf, DataStructures
function funcfreqs(expr::Expr)
cnt = counter(Symbol)
expr.head == :call &&
push!(cnt, expr.args[1])
for e in expr.args
e isa Expr && merge!(cnt, funcfreqs(e))
end
return cnt
end
function parseall(str::AbstractString)
exs = Any[]
pos = start(str)
while !done(str, pos)
ex, pos = parse(str, pos) # returns next starting point as well as expr
ex.head == :toplevel ? append!(exs, ex.args) : push!(exs, ex)
end
if isempty(exs)
throw(ParseError("end of input"))
elseif length(exs) == 1
return exs[1]
else
return Expr(:block, exs...)
end
end
freqs = readstring("src/Function_frequency.jl") |> parseall |> funcfreqs
for (v, f) in freqs
@printf("%10s → %i\n", v, f)
end</lang>
- Output:
append! → 1
isa → 1
start → 1
push! → 2
|> → 2
funcfreqs → 2
parseall → 1
ParseError → 1
! → 1
length → 1
throw → 1
parse → 1
readstring → 1
== → 3
counter → 1
Expr → 1
merge! → 1
isempty → 1
done → 1
LiveCode
Initially based on Listing all the handlers in a script <lang LiveCode>function handlerNames pScript
put pScript into pScriptCopy
filter pScript with regex pattern "^(on|function).*"
-- add in the built-in commands & functions
put the commandNames & the functionnames into cmdfunc
repeat for each line builtin in cmdfunc
put 0 into handlers[builtin]
end repeat
-- add user defined handlers, remove this section of you do not want your own functions included
repeat with x = 1 to the number of lines of pScript
put word 2 of line x of pScript into handlername
put 0 into handlers[handlername]
end repeat
-- count handlers used
repeat with x = 1 to the number of lines of pScriptCopy
repeat for each key k in handlers
if k is among the tokens of line x of pScriptCopy then
add 1 to handlers[k]
end if
end repeat
end repeat
combine handlers using cr and space
sort lines of handlers descending by word 2 of each
put line 1 to 10 of handlers into handlers
return handlers
end handlerNames</lang>
To use<lang LiveCode>put handlerNames(the script of this stack & cr & the script of this card & cr & the script of me)</lang>
Sample output<lang LiveCode>if 8 put 8 return 8 function 7 factorialacc 4 -- user def function for other rosetta task factorialr 3 -- user def function for other rosetta task handlerNames 3 factorial 2 -- user def function for other rosetta task factorialit 2 -- user def function for other rosetta task mouseUp 2</lang>
Mathematica / Wolfram Language
<lang Mathematica>programCount[fn_] := Reverse[If[Length[#] > 10, Take[#, -10], #] &[SortBy[Tally[Cases[DownValues[fn], s_Symbol, \[Infinity], Heads -> True]], Last]]]</lang>
- Output:
The output of applying this program to itself...
programCount[programCount]
{{Slot, 3}, {Pattern, 2}, {fn, 2}, {Blank, 2}, {\[Infinity], 1}, {True, 1}, {Tally, 1}, {Take, 1}, {Symbol, 1}, {SortBy, 1}}
Perl
We leverage the PPI::Tokenizer module. <lang perl>use PPI::Tokenizer; my $Tokenizer = PPI::Tokenizer->new( '/path/to/your/script.pl' ); my %counts; while (my $token = $Tokenizer->get_token) {
# We consider all Perl identifiers. The following regex is close enough.
if ($token =~ /\A[\$\@\%*[:alpha:]]/) {
$counts{$token}++;
}
} my @desc_by_occurrence =
sort {$counts{$b} <=> $counts{$a} || $a cmp $b}
keys(%counts);
my @top_ten_by_occurrence = @desc_by_occurrence[0 .. 9]; foreach my $token (@top_ten_by_occurrence) {
print $counts{$token}, "\t", $token, "\n";
}</lang>
- Output:
When run on itself:
6 $token 6 my 4 $counts 2 $Tokenizer 2 $a 2 $b 2 %counts 2 @desc_by_occurrence 2 @top_ten_by_occurrence 2 PPI::Tokenizer
Phix
As Phix is self hosted, we can modify the compiler (or a copy of it) directly for this task.
Add the line shown to procedure Call() in pmain.e, after the else on line 4938 (at the time of writing)
else -- rType=FUNC|TYPE
log_function_call(rtnNo)
(I may have been a bit too literal about "function" here, specifically "not procedure")
Now create our test.exw program, which wraps the entire compiler: <lang Phix>constant func_log = new_dict(),
func_freq = new_dict()
global procedure log_function_call(integer rtnNo)
integer node = getd_index(rtnNo,func_log) setd(rtnNo,iff(node=NULL?1:getd_by_index(node,func_log)+1),func_log)
end procedure
include p.exw -- the phix compiler, full source
-- invert the dictionary, then print top ten
integer count = 0 function visitor(object key, integer data, integer user_data)
if user_data=1 then -- invert
setd({data,key},0,func_freq)
else
key[2] = symtab[key[2]][S_Name]
?key
count += 1
if count>10 then return 0 end if -- cease traversal
end if
return 1
end function constant r_visitor = routine_id("visitor")
rebuild_callback() -- (convert ternary tree indexes to readable names)
traverse_dict(r_visitor,1,func_log) -- invert traverse_dict(r_visitor,2,func_freq,rev:=true) -- top 10</lang> Invoke using "p test -norun test" (note you can omit the ".exw" part of "test.exw")
- Output:
{1253,"length"}
{655,"and_bits"}
{354,"append"}
{308,"find"}
{163,"or_bits"}
{158,"repeat"}
{154,"SetField"}
{136,"sprintf"}
{119,"equal"}
{105,"sequence"}
{90,"platform"}
Notes:
The log_function call is passed an index into the symbol table.
For performance reasons the compiler uses integer indexes, so we need to invoke
rebuild_callback() to replace them with human-readable names in the symbol table.
For more details of S_Name and other constants/contents of the symbol table, see pglobals.e
The compiler (p.exe) interprets test.exw(+p.exw) which compiles a third copy of itself under -norun.
Notice that it is not necessary to compile the compiler (using p -c p) to test changes in it, and in
fact weeks or months of work on the compiler often happens purely in interpreter mode, between actually
creating a new executable.
If, instead, you want to know how many times a function is called at run-time, just add "with profile"
to the source and it will create a ex.pro listing which tells you.
Lastly, remember to remove/comment out that log_function_call() in pmain.e
PicoLisp
<lang PicoLisp>(let Freq NIL
(for "L" (filter pair (extract getd (all)))
(for "F"
(filter atom
(fish '((X) (or (circ? X) (getd X)))
"L" ) )
(accu 'Freq "F" 1) ) )
(for X (head 10 (flip (by cdr sort Freq)))
(tab (-7 4) (car X) (cdr X)) ) )</lang>
Output, for the system in debug mode plus the above code:
quote 310 car 236 cdr 181 setq 148 let 136 if 127 and 124 cons 110 cadr 80 or 76
If the condition in the 5th line (getd X) is replaced with (sym? X), then all symbols are counted, and the output is
X 566 quote 310 car 236 cdr 181 C 160 N 157 L 155 Lst 152 setq 148 T 144
And if it is replaced with (num? X), it is
1 71 0 38 2 27 3 17 7 9 -1 9 100 8 48 6 43 6 12 6
Python
This code parses a Python source file using the built-in ast module and counts simple function calls; it won't process method calls or cases when you call the result of an expression. Also, since constructors are invoked by calling the class, constructor calls are counted as well. <lang python>import ast
class CallCountingVisitor(ast.NodeVisitor):
def __init__(self):
self.calls = {}
def visit_Call(self, node):
if isinstance(node.func, ast.Name):
fun_name = node.func.id
call_count = self.calls.get(fun_name, 0)
self.calls[fun_name] = call_count + 1
self.generic_visit(node)
filename = input('Enter a filename to parse: ') with open(filename, encoding='utf-8') as f:
contents = f.read()
root = ast.parse(contents, filename=filename) #NOTE: this will throw a SyntaxError if the file isn't valid Python code visitor = CallCountingVisitor() visitor.visit(root) top10 = sorted(visitor.calls.items(), key=lambda x: x[1], reverse=True)[:10] for name, count in top10:
print(name,'called',count,'times')
</lang>
The result of running the program on the ftplib module of Python 3.2:
Enter a filename to parse: c:\Python32\Lib\ftplib.py error_reply called 10 times print called 10 times error_proto called 8 times callback called 8 times int called 7 times repr called 7 times isinstance called 6 times len called 6 times ValueError called 3 times parse257 called 2 times
Racket
<lang racket>
- lang racket
(require math) (define in (open-input-file "function-frequency.rkt")) (void (read-language in)) (define s-exprs (for/list ([s (in-port read in)]) s)) (define symbols (filter symbol? (flatten s-exprs))) (define counts (sort (hash->list (samples->hash symbols)) >= #:key cdr)) (take counts (min 10 (length counts))) </lang> Output: <lang racket> '((define . 4)
(counts . 3) (s-exprs . 2) (s . 2) (symbols . 2) (a-program . 2) (filter . 1) (hash->list . 1) (in-port . 1) (sort . 1))
</lang>
Raku
(formerly Perl 6) Here we just examine the ast of the Raku compiler (which is written in Raku) to look for function calls. <lang perl6>my $text = qqx[raku --target=ast @*ARGS[]]; my %fun; for $text.lines {
%fun{$0}++ if / '(call &' (.*?) ')' /
}
for %fun.invert.sort.reverse[^10] { .value.say }</lang>
- Output:
Here we run it on the strand sort RC entry. Note how Raku considers various operators to really be function calls underneath.
$ ./morefun strand pop postcircumfix:<[ ]> unshift succeed splice prefix:<-> push infix:<,> infix:<..> infix:<->
REXX
version 1
This program counts statically. It lacks, however, treatment of comments and literal strings. <lang rexx>fid='pgm.rex' cnt.=0 funl= Do While lines(fid)>0
l=linein(fid)
Do Until p=0
p=pos('(',l)
If p>0 Then Do
do i=p-1 To 1 By -1 While is_tc(substr(l,i,1))
End
fn=substr(l,i+1,p-i-1)
If fn<> Then
Call store fn
l=substr(l,p+1)
End
End
End
Do While funl<>
Parse Var funl fn funl Say right(cnt.fn,3) fn End
Exit x=a(3)+bbbbb(5,c(555)) special=date('S') 'DATE'() "date"() is_tc: abc='abcdefghijklmnopqrstuvwxyz' Return pos(arg(1),abc||translate(abc)'1234567890_"')>0
store: Parse Arg fun cnt.fun=cnt.fun+1 If cnt.fun=1 Then
funl=funl fun
Return</lang>
- Output:
1 lines 1 linein 2 pos 1 ' 1 is_tc 3 substr 1 right 1 a 1 bbbbb 1 c 1 date 1 'DATE' 1 "date" 1 arg 1 translate
version 2
This program counts statically. Contents of comments and literal strings are not analyzed. Neither are function invocations via CALL. <lang rexx>/* REXX ****************************************** Version 11.12.2015 **
- Rexx Tokenizer to find function invocations
- -----------------------------------------------------------------------
- Tokenization remembers the following for each token
- t.i text of token
- t.i.0t type of token: Cx/V/K/N/O/S/L
- comment/variable/keyword/constant/operator/string/label
- t.i.0il line of token in the input
- t.i.0ic col of token in the input
- t.i.0prev index of token starting previous instruction
- t.i.0ol line of token in the output
- t.i.0oc col of token in the output
- ---------------------------------------------------------------------*/
Call time 'R'
Parse Upper Arg fid '(' options
If fid='?' Then Do
Say 'Tokenike a REXX proram and list the function invocations found'
Say ' which are of the form symbol(... or string(...'
Say ' (the left parenthesis must immediately follow the symbol'
Say ' or literal string.)'
Say 'Syntax:'
Say ' TKZ pgm < ( <Debug> <Tokens> >'
Exit
End
g.=0
Call init /* Initialize constants etc. */
g.0cont='01'x
g.0breakc='02'x
cnt.=0
Call readin /* Read input file into l.* */
Call tokenize /* Tokenize the input */
tk=
Call process_tokens
g.0fun_list=wordsort(g.0fun_list)
Do While g.0fun_list>
Parse Var g.0fun_list fun g.0fun_list
Say right(cnt.fun,3) fun
End
Say time('E') 'seconds elapsed for' t.0 'tokens in' g.0lines 'lines.'
Exit
init: /***********************************************************************
- Initialize constants etc.
- /
g.= g.0debug=0 /* set debug off by default */
fid=strip(fid) If fid= Then /* no file specified */ Exit exit(12 'no input file specified') Parse Var fid fn '.'
os=options /* options specified on command */
g.0debug=0 /* turn off debug output */
g.0tokens=0 /* No token file */
Do While os<> /* process them individually */
Parse Upper Var os o os /* pick one */
Select
When abbrev('DEBUG',o,1) Then /* Debug specified */
g.0debug=1 /* turn on debug output */
When abbrev('TOKENS',o,1) Then /* Write a file with tokens */
g.0tokens=1
Otherwise /* anything else */
Say 'Unknown option:' o /* tell the user and ignore it */
End
End
If g.0debug Then Do g.0dbg=fn'.dbg'; '@erase' g.0dbg End If g.0tokens Then Do g.0tkf=fn'.tok'; '@erase' g.0tkf End
/***********************************************************************
- Language specifics
- /
g.0special='+-*/%";:<>^\=|,()& '/* special characters */
/* chars that may start a var */
g.0a='abcdefghijklmnopqrstuvwxyz'||,
'ABCDEFGHIJKLMNOPQRSTUVWXYZ@#$!?_'
g.0n='1234567890' /* numeric characters */
g.0vc=g.0a||g.0n||'.' /* var-character */
/* multi-character operators */
g.0opx='&& ** // << <<= <= <> == >< >= >> >>=',
'^< ^<< ^= ^== ^> ^>> \< \<< \= \== \> \>> ||'
t.= /* token list */ Return
readin: /***********************************************************************
- Read the file to be formatted
- /
lc=
i=0
g.0lines=0
Do While lines(fid)<>0
li=linein(fid)
g.0lines=g.0lines+1
If i>0 Then
lc=strip(l.i,'T')
If right(lc,1)=',' Then Do
l.i=left(lc,length(lc)-1) li
End
Else Do
i=i+1
l.i=li
End
End
l.0=i
Call lineout fid
t=l.0+1
l.t=g.0eof /* add a stopper at program end */
l.0=t /* adjust number of lines */
g.0il=t /* remember end of program */
Return
tokenize: /***********************************************************************
- First perform tokenization
- Input: l.* Program text
- Output: t.* Token list
- t.0t.i token type CA CB CC C comment begin/middle/end
- S string
- O operator (special character)
- V variable symbol
- N constant
- X end of text
- Note: special characters are treated as separate tokens
- /
li=0 /* line index */
ti=0 /* token index */
Do While li<l.0 /* as long as there is more input */
li=li+1 /* index of next line */
l=l.li /* next line to be processed */
g.0newline=1
g.0cc=0 /* current column */
Call dsp l.li /* debug output */
If l= Then /* empty line */
Call addtoken '/*--*/','C' /* preserve with special token */
Do While l<> /* work through the line */
nbc=verify(l,' ') /* first non-blank column */
g.0cc=g.0cc+nbc /* advance to this */
If g.0newline= Then Do
If t.ti.0ic= Then
t.ti.0ic=0
If g.0cc=t.ti.0ic+length(t.ti) Then Do
tj=ti+1
t.tj.0ad=1
End
End
l=substr(l,nbc) /* and continue with rest of line */
Parse Var l c +1 l 1 c2 +2 /* get character(s) */
g.0tb=g.0cc /* remember where token starts */
Select /* take a decision */
When c2='/*' Then /* comment starts here */
Call comment /* process comment */
When pos(c,"')>0 Then /* literal string starts here */
Call string c /* process literal string */
Otherwise /* neither comment nor literal */
Call token /* get other token */
End /* cmt, string, or token done */
End /* end of loop over line */
End /* end of loop over program */
t.0=ti /* store number of tokens */
Call dsp ti 'tokens' l.0 'lines'
Return
comment: /***********************************************************************
- Parse a comment
- Nested comments are supported
- /
cbeg=t.ti.0il
l=substr(l,2) /* continue after slash-asterisk */
g.0cc=g.0cc+1 /* update current char position */
t='/*' /* token so far */
incmt=1 /* indicate "within a comment" */
Do Until incmt=0 /* loop until done */
bc=pos('/*',l) /* next begin comment, if any */
ec=pos('*/',l) /* next end comment, if any */
Select /* decide */
When bc>0 &, /* begin-comment found */
(ec=0 | bc<ec) Then Do /* and no end-comment or later */
t=t||left(l,bc+1) /* add this all to token */
incmt=incmt+1 /* increment comment nest-depth */
l=substr(l,bc+2) /* continue after slash-asterisk */
g.0cc=g.0cc+bc+1 /* update current char position */
End
When ec>0 Then Do /* end-comment found */
t=t||left(l,ec+1) /* add all to token */
incmt=incmt-1 /* decrement nesting */
l=substr(l,ec+2) /* continue after asterisk-slash */
g.0cc=g.0cc+ec+1 /* update current char position */
End
Otherwise Do /* no further comment bracket */
Call addtoken t||l,ct() /* rest of line to token */
li=li+1 /* proceed to next line */
l=l.li /* contents of next line */
g.0newline=1
If l=g.0eof Then Do
Say 'Comment started in line' cbeg 'is not closed before EOF'
Exit err(58)
End
g.0cc=0 /* current char (none) */
g.0tb=1 /* token (comment) starts here */
End
End
End
Call addtoken t,ct() /* last (or only) comment token */
If pos('*debug*',t)>0 Then g.0debug=1
Return
ct: /***********************************************************************
- Comment type
- /
If incmt>0 Then Do /* within a comment */
If t.ti.0t='CA' |, /* prev. token was start or cont */
t.ti.0t='CB' Then Return 'CB' /* this is continuation */
Else Return 'CA' /* this is start */
End
Else Do /* comment is over */
If t.ti.0t='CA' |, /* prev. token was start or cont */
t.ti.0t='CB' Then Return 'CC' /* this is final part */
Else Return 'C' /* this is just a comment */
End
string: /***********************************************************************
- Parse a string
- take care of '111'B and '123'X
- /
Parse Arg delim /* string delimiter found */
t=delim /* star building the token */
instr=1 /* note we are within a string */
g.0ss=li
Do Until instr=0 /* continue until it is over */
se=pos(delim,l) /* ending delimiter */
If se>0 Then Do /* found */
If substr(l,se+1,1)=delim Then Do /* but it is doubled */
t=t||left(l,se+1) /* so add all so far to token */
l=substr(l,se+2) /* and take rest of line */
g.0cc=g.0cc+se+1 /* and set current character pos */
End
Else Do /* not another one */
instr=0 /* string is done */
t=t||left(l,se) /* add the string data to token */
l=substr(l,se+1) /* take the rest of the line */
g.0cc=g.0cc+se /* and set current character pos */
If pos(translate(left(l,1)),'BX')>0 Then
If pos(substr(l,2,1),g.0vc)=0 Then Do
t=t||left(l,1) /* add the char to the token */
l=substr(l,2) /* take the rest of the line */
g.0cc=g.0cc+1 /* and set current character pos */
End
End
End
Else Do /* not found */
Call addtoken t||l,'S' /* store the token */
g.0lasttoken= /* reset this switch */
li=li+1 /* go on to the next line */
If li>l.0 Then /* there is no next line */
Exit err(60,'string starting in line' g.0ss,
'does not end before end of file')
Else
Say 'string starting at line' g.0ss 'extended over line boundary'
l=l.li /* take contents of the next line */
g.0cc=1 /* current char position */
g.0tb=1 /* ?? */
End
End
Call addtoken t,'S' /* store the token */
Return
token: /***********************************************************************
- Parse a token
- /
IF c=g.0comma & l= Then Do
t=g.0cont
type='O' /* O (for operator - not quite...)*/
End
Else Do
If pos(c,g.0special)>0 Then Do /* a special character */
t=c /* take it as is */
type='O' /* O (for operator - not quite...)*/
End
Else Do /* some other character */
nsp=verify(l,g.0special,'M') /* find delimiting character */
If nsp>0 Then Do /* some character found */
t=c||left(l,nsp-1) /* take all up to this character */
l=substr(l,nsp) /* and continue from there */
End
Else Do /* none found */
t=c||l /* add rest of line to token */
l= /* and all is used up */
End
g.0cc=g.0cc+length(t)-1 /* adjust current char position */
If pos(right(t,1),'eE')>0 &, /* consider nxxxE+nn case */
pos(left(l,1),'+-')>0 Then Do
If pos(left(t,1),'.1234567890')>0 Then /* start . or digit */
If pos(substr(l,2,1),'1234567890')>0 Then Do /* dig after+- */
nsp=verify(substr(l,2),g.0special,'M')+1 /* find end */
If nsp>1 Then /* delimiting character found */
exp=substr(l,2,nsp-2) /* exponent (if numeric) */
Else
exp=substr(l,2)
If verify(exp,'0123456789')=0 Then Do
t=t||left(l,1)||exp
l=substr(l,length(exp)+2)
g.0cc=g.0cc+length(exp)+2
End
End
End
Select
When isvar(t) Then /* token qualifies as variable */
type='V'
When isconst(t) Then /* token is a constant symbol */
type='N'
When t=g.0eof Then /* token is end of file indication*/
type='X'
Otherwise Do /* anything else is an error */
Say 'li='li
Say l
Say 'token error'
Trace ?R
Exit err(62,'token' t 'is neither variable nor constant')
End
End
If left(l,1)='(' Then
type=type||'F'
End
End
Call addtoken t,type /* store the token */
Return
addtoken: /***********************************************************************
- Add a token to the token list
- /
Parse Arg t,type /* token and its type */
If type='O' Then Do /* operator (special character) */
If pos(t,'><=&|/*')>0 Then Do /* char for composite operator */
If wordpos(t.ti||t,g.0opx)>0 Then Do /* composite operator */
t.ti=t.ti||t /* use concatenation */
/* does not handle =/**/= */
t= /* we are done */
Return
End
End
End
If type='CC' & t='*/' Then Do /* The special case for SPA */ Return End
ti=ti+1 /* increment index */
t.ti=t /* store token's value */
t.ti.0t=left(type,1) /* and its type */
t.ti.0nl=g.0newline /* token starts a new line */
g.0newline= /* reset new line switch */
If t.ti.0t='C' Then Do
t.ti.0t=type
If left(t.ti,3)='/* ' &,
right(t.ti,3)=' */' Then
t.ti='/*' strip(substr(t.ti,4,length(t.ti)-6)) '*/'
End
t.ti.0f=substr(type,2,1) /* 'F' if possibly a function */
Call setpos ti li g.0tb /* and its position */
If left(type,1)='C' Then /* ??? */
If left(t.ti,2)<>'/*' Then Do
ts=strip(t.ti,'L')
t.ti.0oc=t.ti.0oc+length(t.ti)-length(ts)
t.ti=ts
End
If t.ti.0ol= Then t.ti.0ol=li
If t.ti.0oc= Then t.ti.0oc=0
t.ti.0il=t.ti.0ol /* and its position */
t.ti.0ic=t.ti.0oc /* and its position */
Call dsp ti t.ti t.ti.0il'/'t.ti.0ic '->' t.ti.0ol'/'t.ti.0oc
t= /* reset token variable */
Return
lookback: /***********************************************************************
- Look back if...
- /
Do i_=ti To 1 By -1
Select
When left(t.i_.0t,1)='C' Then Nop
When t.i_.0used<>1 &,
(t.i_=g.0comma |,
t.i_=g.0cont) Then Do
t.i_.0used=1
t.i_=g.0cont
Return '0'
End
Otherwise
Return '1'
End
End
Return '1'
isvar: /***********************************************************************
- Determine if a string qualifies as variable name
- /
Parse Arg a_ +1 b_ res=(pos(a_,g.0a)>0) &, (verify(b_,g.0a||g.0n||'.')=0) Return res
isconst: /***********************************************************************
- Determine if a string qualifies as constant
- /
Parse Arg a_ res=(verify(a_,g.0a||g.0n||'.+-')=0) /* ??? */ Return res
setpos:
Parse Arg seti sol soc
setz='setpos:' t.seti t.seti.0ol'/'t.seti.0oc '-->',
sol'/'soc '('sigl')'
Call dsp setz
t.seti.0ol=sol
t.seti.0oc=soc
Return
process_tokens: /***********************************************************************
- Process the token list
- /
Do i=1 To t.0
If g.0tokens Then
Call lineout g.0tkf,right(i,4) right(t.i.0il,3)'.'left(t.i.0ic,3),
right(t.i.0ol,3)'.'left(t.i.0oc,3),
left(t.i.0t,2) left(t.i,25)
If t.i='(' Then Do
j=i-1
If t.j.0ol=t.i.0il & ,
t.j.0oc+length(t.j)=t.i.0ic &,
pos(t.j.0t,'VS')>0 Then
Call store_f t.j
End
End
If g.0tokens Then
Call lineout g.0tkf
Return
store_f:
Parse Arg funct If wordpos(funct,g.0fun_list)=0 then g.0fun_list=g.0fun_list funct cnt.funct=cnt.funct+1 Return
dsp: /***********************************************************************
- Record (and display) a debug line
- /
Parse Arg ol_.1 If g.0debug>0 Then Call lineout g.0dbg,ol_.1 If g.0debug>1 Then Say ol_.1 Return
wordsort: Procedure /**********************************************************************
- Sort the list of words supplied as argument. Return the sorted list
- /
Parse Arg wl
wa.=
wa.0=0
Do While wl<>
Parse Var wl w wl
Do i=1 To wa.0
If wa.i>w Then Leave
End
If i<=wa.0 Then Do
Do j=wa.0 To i By -1
ii=j+1
wa.ii=wa.j
End
End
wa.i=w
wa.0=wa.0+1
End
swl=
Do i=1 To wa.0
swl=swl wa.i
End
Return strip(swl)
err: /***********************************************************************
- Diagnostic error exit
- /
Parse Arg errnum, errtxt
Say 'err:' errnum errtxt
If t.ti.0il>g.0il Then
Say 'Error' arg(1) 'at end of file'
Else Do
Say 'Error' arg(1) 'around line' t.ti.0il', column' t.ti.0ic
_=t.ti.0il
Say l._
Say copies(' ',t.ti.0ic-1)'|'
End
If errtxt<> Then Say ' 'errtxt
Exit 12</lang>
- Output:
Result for the above program.
2 abbrev 2 arg 1 copies 2 ct 3 err 1 exit 1 isconst 1 isvar 21 left 9 length 1 linein 1 lines 15 pos 7 right 5 strip 17 substr 1 time 1 translate 6 verify 2 wordpos 1 wordsort 0.093000 seconds elapsed for 2200 tokens in 510 lines.
Sidef
Sidef provides full access to its parser, allowing us to inspect all the declarations within a program. <lang ruby>func foo { } func bar { }
foo(); foo(); foo() bar(); bar();
var data = Perl.to_sidef(Parser{:vars}{:main}).flatten
data.sort_by { |v| -v{:count} }.first(10).each { |entry|
if (entry{:type} == :func) {
say ("Function `#{entry{:name}}` (declared at line",
" #{entry{:line}}) is used #{entry{:count}} times")
}
}</lang>
- Output:
Function `foo` (declared at line 1) is used 3 times Function `bar` (declared at line 2) is used 2 times
Smalltalk
This parses all classes of all loaded packages/libraries (takes a few seconds). From the code, it should be obvious how to restrict the search to packages, libraries, classes or individual methods.
<lang smalltalk>bagOfCalls := Bag new. Smalltalk allClassesDo:[:cls |
cls instAndClassMethodsDo:[:mthd |
bagOfCalls addAll:mthd messagesSent
].
]. (bagOfCalls sortedCounts to:10) do:[:assoc |
Stdout printCR: e'method {assoc value} is called {assoc key} times.'
].</lang> note: messagesSent calls the parser for an AST and enumerates the parse nodes; it does not know, which get inlined and which end up being called actually (in fact, most of the one's below are probably inlined).
The results below were printed in scripting mode, where only a small part of the system is actually preloaded.
- Output:
method ifTrue:is called 19805 times. method notNil is called 8731 times. method isNil is called 8616 times. method ifTrue:ifFalse: is called 6833 times. method == is called 6655 times. method new is called 6088 times. method ifFalse: is called 4640 times. method and: is called 4389 times. method size is called 3800 times. method + is called 3737 times.
Tcl
<lang tcl>package require Tcl 8.6
proc examine {filename} {
global cmds
set RE "(?:^|\[\[\{\])\[\\w:.\]+"
set f [open $filename]
while {[gets $f line] >= 0} {
set line [string trim $line] if {$line eq "" || [string match "#*" $line]} { continue } foreach cmd [regexp -all -inline $RE $line] { incr cmds([string trim $cmd "\{\["]) }
} close $f
}
- Parse each file on the command line
foreach filename $argv {
examine $filename
}
- Get the command list in order of frequency
set cmdinfo [lsort -stride 2 -index 1 -integer -decreasing [array get cmds]]
- Print the top 10 (two list items per entry, so 0-19, not 0-9)
foreach {cmd count} [lrange $cmdinfo 0 19] {
puts [format "%-20s%d" $cmd $count]
}</lang> Sample run (note that the commands found are all standard Tcl commands; they're just commands so it is natural to expect them to be found):
bash$ tclsh8.6 RosettaCode/cmdfreq.tcl RosettaCode/*.tcl set 2374 expr 846 if 775 puts 558 return 553 proc 549 incr 485 foreach 432 lindex 406 lappend 351