commit 90ac09c942c36b7bf8c508b37be649afea2e9090
parent a36f1c294c2677674520216a17236beaceb89de7
Author: Santtu Lakkala <inz@inz.fi>
Date: Mon, 21 Feb 2022 16:19:15 +0200
[st][patch][copyurl] Slight update and multiline version
The previous version of my copyurl patch worked incorrectly due to
relying on the snapping selection; use double selextend() instead.
Also a new version that uses a (reverse) DFA to detect links without
copying to buffer, enabling the detection of multiple line urls.
Diffstat:
4 files changed, 300 insertions(+), 137 deletions(-)
diff --git a/st.suckless.org/patches/copyurl/index.md b/st.suckless.org/patches/copyurl/index.md
@@ -22,7 +22,12 @@ Following patches also highlight the selected urls:
* [st-copyurl-20190202-3be4cf1.diff](st-copyurl-20190202-3be4cf1.diff)
* [st-copyurl-0.8.4.diff](st-copyurl-0.8.4.diff)
-* [st-copyurl-20220217-0.8.5.diff](st-copyurl-20220217-0.8.5.diff)
+* [st-copyurl-20220221-0.8.5.diff](st-copyurl-20220221-0.8.5.diff)
+
+Following patches also allow urls that span multiple lines:
+
+* [st-copyurl-multiline-20220221-0.8.5.diff](st-copyurl-multiline-20220221-0.8.5.diff)
+
Authors
-------
@@ -31,3 +36,4 @@ Authors
* David Phillips - <david@sighup.nz> (git port)
* Matthew Parnell - <matt@parnmatt.co.uk> (0.7, git ports)
* Michael Buch - <michaelbuch12@gmail.com> (0.8+ git port, url highlighting)
+* Santtu Lakkala - <inz@inz.fi> (multiline url support)
diff --git a/st.suckless.org/patches/copyurl/st-copyurl-20220217-0.8.5.diff b/st.suckless.org/patches/copyurl/st-copyurl-20220217-0.8.5.diff
@@ -1,136 +0,0 @@
-From eacbcf7de84e81a27759d086ad1a5ea9e41f9894 Mon Sep 17 00:00:00 2001
-From: Santtu Lakkala <inz@inz.fi>
-Date: Wed, 16 Feb 2022 20:34:20 +0200
-Subject: [PATCH] Loop through urls on screen and copy to clipboard
-
-Based on the previous highlighting patches, slightly simplified and
-fixes graphical issues with mixed copyurl and selection.
----
- config.def.h | 1 +
- st.c | 81 ++++++++++++++++++++++++++++++++++++++++++++++++++++
- st.h | 1 +
- 3 files changed, 83 insertions(+)
-
-diff --git a/config.def.h b/config.def.h
-index 91ab8ca..3f365c7 100644
---- a/config.def.h
-+++ b/config.def.h
-@@ -201,6 +201,7 @@ static Shortcut shortcuts[] = {
- { TERMMOD, XK_Y, selpaste, {.i = 0} },
- { ShiftMask, XK_Insert, selpaste, {.i = 0} },
- { TERMMOD, XK_Num_Lock, numlock, {.i = 0} },
-+ { MODKEY, XK_l, copyurl, {.i = 0} },
- };
-
- /*
-diff --git a/st.c b/st.c
-index 51049ba..931b29f 100644
---- a/st.c
-+++ b/st.c
-@@ -200,6 +200,7 @@ static void tdefutf8(char);
- static int32_t tdefcolor(const int *, int *, int);
- static void tdeftran(char);
- static void tstrsequence(uchar);
-+static const char *findlastany(const char *, const char**, size_t);
-
- static void drawregion(int, int, int, int);
-
-@@ -2688,3 +2689,83 @@ redraw(void)
- tfulldirt();
- draw();
- }
-+
-+const char *
-+findlastany(const char *str, const char**find, size_t len)
-+{
-+ const char *found = NULL;
-+ int i = 0;
-+
-+ for (found = str + strlen(str) - 1; found >= str; --found) {
-+ for(i = 0; i < len; i++) {
-+ if (strncmp(found, find[i], strlen(find[i])) == 0) {
-+ return found;
-+ }
-+ }
-+ }
-+
-+ return NULL;
-+}
-+
-+/*
-+** Select and copy the previous url on screen (do nothing if there's no url).
-+**
-+** FIXME: doesn't handle urls that span multiple lines; will need to add support
-+** for multiline "getsel()" first
-+*/
-+void
-+copyurl(const Arg *arg) {
-+ /* () and [] can appear in urls, but excluding them here will reduce false
-+ * positives when figuring out where a given url ends.
-+ */
-+ static const char URLCHARS[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
-+ "abcdefghijklmnopqrstuvwxyz"
-+ "0123456789-._~:/?#@!$&'*+,;=%";
-+
-+ static const char* URLSTRINGS[] = {"http://", "https://"};
-+
-+ int row = 0, /* row of current URL */
-+ col = 0, /* column of current URL start */
-+ colend = 0, /* column of last occurrence */
-+ passes = 0; /* how many rows have been scanned */
-+
-+ char linestr[term.col + 1];
-+ const char *c = NULL,
-+ *match = NULL;
-+
-+ row = (sel.ob.x >= 0 && sel.nb.y > 0) ? sel.nb.y : term.bot;
-+ LIMIT(row, term.top, term.bot);
-+
-+ colend = (sel.ob.x >= 0 && sel.nb.y > 0) ? sel.nb.x : term.col;
-+ LIMIT(colend, 0, term.col);
-+
-+ /*
-+ ** Scan from (term.row - 1,term.col - 1) to (0,0) and find
-+ ** next occurrance of a URL
-+ */
-+ for (passes = 0; passes < term.row; passes++) {
-+ /* Read in each column of every row until
-+ ** we hit previous occurrence of URL
-+ */
-+ for (col = 0; col < colend; ++col)
-+ linestr[col] = term.line[row][col].u < 128 ? term.line[row][col].u : ' ';
-+ linestr[col] = '\0';
-+
-+ if ((match = findlastany(linestr, URLSTRINGS,
-+ sizeof(URLSTRINGS)/sizeof(URLSTRINGS[0]))))
-+ break;
-+
-+ if (--row < 0)
-+ row = term.row - 1;
-+
-+ colend = term.col;
-+ };
-+
-+ if (match) {
-+ size_t l = strspn(match, URLCHARS);
-+ selstart(match - linestr, row, 1);
-+ selextend(match - linestr + l - 1, row, 1, SEL_REGULAR);
-+ xsetsel(getsel());
-+ xclipcopy();
-+ }
-+}
-diff --git a/st.h b/st.h
-index 519b9bd..0458005 100644
---- a/st.h
-+++ b/st.h
-@@ -85,6 +85,7 @@ void printscreen(const Arg *);
- void printsel(const Arg *);
- void sendbreak(const Arg *);
- void toggleprinter(const Arg *);
-+void copyurl(const Arg *);
-
- int tattrset(int);
- void tnew(int, int);
---
-2.32.0
-
diff --git a/st.suckless.org/patches/copyurl/st-copyurl-20220221-0.8.5.diff b/st.suckless.org/patches/copyurl/st-copyurl-20220221-0.8.5.diff
@@ -0,0 +1,137 @@
+From 897c3958d01d8df80ebf1666b972b8a658b419ba Mon Sep 17 00:00:00 2001
+From: Santtu Lakkala <inz@inz.fi>
+Date: Wed, 16 Feb 2022 20:34:20 +0200
+Subject: [PATCH] Loop through urls on screen and copy to clipboard
+
+Based on the previous highlighting patches, slightly simplified and
+fixes graphical issues with mixed copyurl and selection.
+---
+ config.def.h | 1 +
+ st.c | 82 ++++++++++++++++++++++++++++++++++++++++++++++++++++
+ st.h | 1 +
+ 3 files changed, 84 insertions(+)
+
+diff --git a/config.def.h b/config.def.h
+index 91ab8ca..3f365c7 100644
+--- a/config.def.h
++++ b/config.def.h
+@@ -201,6 +201,7 @@ static Shortcut shortcuts[] = {
+ { TERMMOD, XK_Y, selpaste, {.i = 0} },
+ { ShiftMask, XK_Insert, selpaste, {.i = 0} },
+ { TERMMOD, XK_Num_Lock, numlock, {.i = 0} },
++ { MODKEY, XK_l, copyurl, {.i = 0} },
+ };
+
+ /*
+diff --git a/st.c b/st.c
+index 51049ba..bf3d81a 100644
+--- a/st.c
++++ b/st.c
+@@ -200,6 +200,7 @@ static void tdefutf8(char);
+ static int32_t tdefcolor(const int *, int *, int);
+ static void tdeftran(char);
+ static void tstrsequence(uchar);
++static const char *findlastany(const char *, const char**, size_t);
+
+ static void drawregion(int, int, int, int);
+
+@@ -2688,3 +2689,84 @@ redraw(void)
+ tfulldirt();
+ draw();
+ }
++
++const char *
++findlastany(const char *str, const char**find, size_t len)
++{
++ const char *found = NULL;
++ int i = 0;
++
++ for (found = str + strlen(str) - 1; found >= str; --found) {
++ for(i = 0; i < len; i++) {
++ if (strncmp(found, find[i], strlen(find[i])) == 0) {
++ return found;
++ }
++ }
++ }
++
++ return NULL;
++}
++
++/*
++** Select and copy the previous url on screen (do nothing if there's no url).
++**
++** FIXME: doesn't handle urls that span multiple lines; will need to add support
++** for multiline "getsel()" first
++*/
++void
++copyurl(const Arg *arg) {
++ /* () and [] can appear in urls, but excluding them here will reduce false
++ * positives when figuring out where a given url ends.
++ */
++ static const char URLCHARS[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
++ "abcdefghijklmnopqrstuvwxyz"
++ "0123456789-._~:/?#@!$&'*+,;=%";
++
++ static const char* URLSTRINGS[] = {"http://", "https://"};
++
++ int row = 0, /* row of current URL */
++ col = 0, /* column of current URL start */
++ colend = 0, /* column of last occurrence */
++ passes = 0; /* how many rows have been scanned */
++
++ char linestr[term.col + 1];
++ const char *c = NULL,
++ *match = NULL;
++
++ row = (sel.ob.x >= 0 && sel.nb.y > 0) ? sel.nb.y : term.bot;
++ LIMIT(row, term.top, term.bot);
++
++ colend = (sel.ob.x >= 0 && sel.nb.y > 0) ? sel.nb.x : term.col;
++ LIMIT(colend, 0, term.col);
++
++ /*
++ ** Scan from (term.row - 1,term.col - 1) to (0,0) and find
++ ** next occurrance of a URL
++ */
++ for (passes = 0; passes < term.row; passes++) {
++ /* Read in each column of every row until
++ ** we hit previous occurrence of URL
++ */
++ for (col = 0; col < colend; ++col)
++ linestr[col] = term.line[row][col].u < 128 ? term.line[row][col].u : ' ';
++ linestr[col] = '\0';
++
++ if ((match = findlastany(linestr, URLSTRINGS,
++ sizeof(URLSTRINGS)/sizeof(URLSTRINGS[0]))))
++ break;
++
++ if (--row < 0)
++ row = term.row - 1;
++
++ colend = term.col;
++ };
++
++ if (match) {
++ size_t l = strspn(match, URLCHARS);
++ selstart(match - linestr, row, 0);
++ selextend(match - linestr + l - 1, row, SEL_REGULAR, 0);
++ selextend(match - linestr + l - 1, row, SEL_REGULAR, 1);
++ xsetsel(getsel());
++ xclipcopy();
++ }
++}
+diff --git a/st.h b/st.h
+index 519b9bd..0458005 100644
+--- a/st.h
++++ b/st.h
+@@ -85,6 +85,7 @@ void printscreen(const Arg *);
+ void printsel(const Arg *);
+ void sendbreak(const Arg *);
+ void toggleprinter(const Arg *);
++void copyurl(const Arg *);
+
+ int tattrset(int);
+ void tnew(int, int);
+--
+2.32.0
+
diff --git a/st.suckless.org/patches/copyurl/st-copyurl-multiline-20220221-0.8.5.diff b/st.suckless.org/patches/copyurl/st-copyurl-multiline-20220221-0.8.5.diff
@@ -0,0 +1,156 @@
+From 30a04d9ecb3998953bdbe42e5617d00d6002869b Mon Sep 17 00:00:00 2001
+From: Santtu Lakkala <inz@inz.fi>
+Date: Wed, 16 Feb 2022 20:34:20 +0200
+Subject: [PATCH] Loop through urls on screen and copy to clipboard
+
+Replace url detection heuristics with a DFA, enabling urls that span
+multiple lines. Also fix the selection not to use snapping so that urls
+are selected exactly.
+---
+ config.def.h | 1 +
+ st.c | 93 ++++++++++++++++++++++++++++++++++++++++++++++++++++
+ st.h | 1 +
+ 3 files changed, 95 insertions(+)
+
+diff --git a/config.def.h b/config.def.h
+index 91ab8ca..3f365c7 100644
+--- a/config.def.h
++++ b/config.def.h
+@@ -201,6 +201,7 @@ static Shortcut shortcuts[] = {
+ { TERMMOD, XK_Y, selpaste, {.i = 0} },
+ { ShiftMask, XK_Insert, selpaste, {.i = 0} },
+ { TERMMOD, XK_Num_Lock, numlock, {.i = 0} },
++ { MODKEY, XK_l, copyurl, {.i = 0} },
+ };
+
+ /*
+diff --git a/st.c b/st.c
+index 51049ba..5b6d919 100644
+--- a/st.c
++++ b/st.c
+@@ -152,6 +152,11 @@ typedef struct {
+ int narg; /* nb of args */
+ } STREscape;
+
++typedef struct {
++ int state;
++ size_t length;
++} URLdfa;
++
+ static void execsh(char *, char **);
+ static void stty(char **);
+ static void sigchld(int);
+@@ -200,6 +205,7 @@ static void tdefutf8(char);
+ static int32_t tdefcolor(const int *, int *, int);
+ static void tdeftran(char);
+ static void tstrsequence(uchar);
++static int daddch(URLdfa *, char);
+
+ static void drawregion(int, int, int, int);
+
+@@ -2688,3 +2694,90 @@ redraw(void)
+ tfulldirt();
+ draw();
+ }
++
++int
++daddch(URLdfa *dfa, char c)
++{
++ /* () and [] can appear in urls, but excluding them here will reduce false
++ * positives when figuring out where a given url ends.
++ */
++ static const char URLCHARS[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
++ "abcdefghijklmnopqrstuvwxyz"
++ "0123456789-._~:/?#@!$&'*+,;=%";
++ static const char RPFX[] = "//:sptth";
++
++ if (!strchr(URLCHARS, c)) {
++ dfa->length = 0;
++ dfa->state = 0;
++
++ return 0;
++ }
++
++ dfa->length++;
++
++ if (dfa->state == 2 && c == '/') {
++ dfa->state = 0;
++ } else if (dfa->state == 3 && c == 'p') {
++ dfa->state++;
++ } else if (c != RPFX[dfa->state]) {
++ dfa->state = 0;
++ return 0;
++ }
++
++ if (dfa->state++ == 7) {
++ dfa->state = 0;
++ return 1;
++ }
++
++ return 0;
++}
++
++/*
++** Select and copy the previous url on screen (do nothing if there's no url).
++*/
++void
++copyurl(const Arg *arg) {
++ int row = 0, /* row of current URL */
++ col = 0, /* column of current URL start */
++ colend = 0, /* column of last occurrence */
++ passes = 0; /* how many rows have been scanned */
++
++ const char *c = NULL,
++ *match = NULL;
++ URLdfa dfa = { 0 };
++
++ row = (sel.ob.x >= 0 && sel.nb.y > 0) ? sel.nb.y : term.bot;
++ LIMIT(row, term.top, term.bot);
++
++ colend = (sel.ob.x >= 0 && sel.nb.y > 0) ? sel.nb.x : term.col;
++ LIMIT(colend, 0, term.col);
++
++ /*
++ ** Scan from (term.row - 1,term.col - 1) to (0,0) and find
++ ** next occurrance of a URL
++ */
++ for (passes = 0; passes < term.row; passes++) {
++ /* Read in each column of every row until
++ ** we hit previous occurrence of URL
++ */
++ for (col = colend; col--;)
++ if (daddch(&dfa, term.line[row][col].u < 128 ? term.line[row][col].u : ' '))
++ break;
++
++ if (col >= 0)
++ break;
++
++ if (--row < 0)
++ row = term.row - 1;
++
++ colend = term.col;
++ }
++
++ if (passes < term.row) {
++ selstart(col, row, 0);
++ selextend((col + dfa.length - 1) % term.col, row + (col + dfa.length - 1) / term.col, SEL_REGULAR, 0);
++ selextend((col + dfa.length - 1) % term.col, row + (col + dfa.length - 1) / term.col, SEL_REGULAR, 1);
++ xsetsel(getsel());
++ xclipcopy();
++ }
++}
+diff --git a/st.h b/st.h
+index 519b9bd..0458005 100644
+--- a/st.h
++++ b/st.h
+@@ -85,6 +85,7 @@ void printscreen(const Arg *);
+ void printsel(const Arg *);
+ void sendbreak(const Arg *);
+ void toggleprinter(const Arg *);
++void copyurl(const Arg *);
+
+ int tattrset(int);
+ void tnew(int, int);
+--
+2.32.0
+
