From 118995a315863961f7b3a8a4509f06279b039b75 Mon Sep 17 00:00:00 2001
From: Holden Rohrer <hr@hrhr.dev>
Date: Wed, 13 May 2020 16:04:07 -0400
Subject: added badroff nbsp and merge

---
 badroff.c | 84 ++++++++++++++++++++++++++++++++++++++++++++++++++++-----------
 1 file changed, 70 insertions(+), 14 deletions(-)

diff --git a/badroff.c b/badroff.c
index a1255d9..6a630ed 100644
--- a/badroff.c
+++ b/badroff.c
@@ -67,6 +67,7 @@ size_t chrnfill(char chr, size_t sz){//fills to first of chr or sz
 
 // typesetting config and commands to change.
 int width = 80;
+char nbsp = 0;
 
 char* center(char* txt){
     size_t len = strlen(txt);
@@ -103,7 +104,7 @@ char* leader(char* txt){
         fin[i] = repeat[i % rptln];
     return fin;
 }
-bool endlineset = false;
+bool endgroup = false;
 llnode* freenext(llnode* node){
     llnode* tmp = node->next;
     node->next = tmp->next;
@@ -125,7 +126,6 @@ void multbreak(sb* buffer, size_t start,
         } else {
             char tmp = str[len];
             str[len] = 0; // truncate str at max length
-            //printf("string %s\n",str);
             insstr(buffer, str); // insert that
             insstr(buffer, "\n");
             str[len] = tmp; // restore string
@@ -138,14 +138,14 @@ llnode* accumlines(size_t* ct){ // returns tail of linked list
     (*ct) = 0;
     while (true){
         tail->next = appendll(tail, line());
-        if (endlineset) break;
+        if (endgroup) break;
         (*ct)++;
         tail = tail->next;
     }
     free(tail->next->str); free(tail->next); //.ELS cmd
     tail->next = head->next; // loop linked list
     free(head);
-    endlineset = false; // prevents false positive on next run
+    endgroup = false; // prevents false positive on next run
     return tail;
 }
 char* lineset(char* txt){
@@ -173,7 +173,6 @@ char* lineset(char* txt){
             tail = tail->next;
         } while (tail != orig);
         if (valid) brk = i;
-        //printf("brk %d\n",brk);
         if (done || (!valid && i-oldbrk >= width)){
             // either all strings are done
             // or it has hit a non-space column and accum enough space
@@ -185,20 +184,61 @@ char* lineset(char* txt){
     char* lines = decompose(buf);
     return lines;
 }
-char* finlineset(char* txt){
+char* emptalloc(void){ // need a null string so it can be freed later
     char* str = malloc(sizeof(char)); str[0] = 0;
-    endlineset = true;
     return str;
 }
+char* fingroup(char* txt){
+    endgroup = true;
+    return emptalloc();
+}
 char* vert(char* txt){
     int ln; sscanf(txt, "%d", &ln);
     return chrmult('\n', ln);
 }
+char* nbrkspc(char* txt){
+    if (txt[0] == '\n') txt[0] = 0;
+    if (txt[1] == '\n') txt[1] = 0;
+    if (txt[0] != 0 && txt[1] != 0)
+        nbsp = txt[1];
+    else if (txt[0] != 0)
+        nbsp = txt[0];
+    else
+        nbsp = 0;
+    return NULL;
+}
+char* merge(char* txt){ // merges until an endgroup
+    // gather lines
+    size_t ct;
+    llnode* head = accumlines(&ct)->next;
+    // count total length (for malloc)
+    size_t len = 0;
+    for (size_t i = 0; i < ct; i++, head = head->next){
+        size_t slen = strlen(head->str)-1;
+        head->str[slen] = 0; // remove the nl character at the end
+        len += slen; // add the (new) length of string
+    }
+    char *str, *pos;
+    pos = str = malloc(sizeof(char)*(len+2));
+    str[len] = '\n'; str[len+1] = 0; // str = "***********\n"
+    for (size_t i = 0; i < ct; i++){
+        wsclean(head->str);
+        size_t slen = strlen(head->str);
+        memcpy(pos, head->str, slen);
+        pos += slen;
+        llnode* tmp = head;
+        head = head->next;
+        free(tmp->str);
+        free(tmp);
+    }
+    return str;
+}
 
 char* cmds[] = // MUST be sorted alphabetically
-    {"CT ", "ELS", "FIL ", "LD ", "LS", "V ", "W "};
+    {"CT ", "EG", "FIL ", "LD ", "LS", "MRG", "NBSP", "V ", "W "};
 char* (*call[])(char* txt) =
-    {center, finlineset, fillline, leader, lineset, vert, setwidth};
+    {center, fingroup, fillline, leader, lineset, merge, nbrkspc,
+    vert, setwidth};
 
 char* cmd(void){
     char* dat = popstrbuf(inbuf, chrfill('\n'));
@@ -237,22 +277,29 @@ void wsclean(char* txt){
         txt[i-disp] = txt[i];
     }
 }
+#define nbspchr (char) 255
+void nbspclean(char* txt){
+    for (; *txt != 0; txt++){
+        if (*txt == nbspchr) *txt = ' ';
+    }
+}
 
 char* wordset(char* txt){
     char* orig = txt; // txt will be manipulated
-    size_t ws = width;
+    size_t ws = 0; // zero means do not cut
     for (size_t i=0; txt[i] != 0; i++){ // turn into a nested loop
         if (txt[i] == ' ' || txt[i] == '\n') ws = i;
         // nl's (multiline input) and spaces are proper linebreaks
-        if (txt[i] == '\n' || (i >= width && ws != i) ){
+        if (txt[i] == '\n' || (ws && i >= width && ws != i) ){
             txt += ws;
             txt[0] = '\n';
-            ws = width;
+            ws = 0;
             i = 0;
         }
     }
 
     wsclean(orig); // remove trailing whitespace
+    nbspclean(orig);
 
     return orig;
 }
@@ -264,6 +311,12 @@ char* norm(void){
 
 // a parser to choose when to typeset and when to run a command
 
+void nbspsub(char* txt){
+    for (; (*txt) != 0; txt++){
+        if (*txt == nbsp) *txt = nbspchr;
+    }
+}
+
 char* line(void){
     size_t sz;
     if ( (sz = fillbuf(2)) == 0)
@@ -271,13 +324,16 @@ char* line(void){
     char* twobytes = peekstrbuf(inbuf, 0, 2); // .., .\n, or ^.?
     if (sz == 1 || twobytes[1] == '\n') return norm();
     if (twobytes[0] == '.') popchrbuf(inbuf);
+    char* out;
     if (twobytes[0] == '.' && twobytes[1] != '.'){
         char* data = cmd();
         if (data) return data;
-        return line();
+        out = line();
     } else {
-        return norm();
+        out = norm();
     }
+    nbspsub(out);
+    return out;
 }
 
 // orchestration
-- 
cgit