7309 - tile: parsing strings into words

2 files changed, 39 insertions, 156 deletions

diff --git a/apps/tile/data.mu b/apps/tile/data.mu
index 6b6d4775..dada0b5b 100644
--- a/apps/tile/data.mu
+++ b/apps/tile/data.mu
@@ -112,31 +112,7 @@ fn create-primitive-functions _self: (addr handle function) {
   var body/eax: (addr line) <- lookup *body-ah
   initialize-line body
   var curr-word-ah/ecx: (addr handle word) <- get body, data
-  # *curr-word = "x"
-  allocate curr-word-ah
-  var tmp/eax: (addr word) <- lookup *curr-word-ah
-  var curr-word/edx: (addr word) <- copy tmp
-  initialize-word-with curr-word, "x"
-  # *curr-word->next = "2"
-  var next-word-ah/ebx: (addr handle word) <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "2"
-  # *curr-word->next->prev = curr-word
-  var prev-word-ah/edi: (addr handle word) <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  # curr-word = curr-word->next
-  curr-word-ah <- copy next-word-ah
-  curr-word <- copy tmp
-  # *curr-word->next = "*"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "*"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  tmp <- lookup *prev-word-ah
+  parse-words "x 2 *", curr-word-ah
   # x 1+ = x 1 +
   var next/esi: (addr handle function) <- get f, next
   allocate next
@@ -152,32 +128,8 @@ fn create-primitive-functions _self: (addr handle function) {
   allocate body-ah
   var body/eax: (addr line) <- lookup *body-ah
   initialize-line body
-  var curr-word-ah/ecx: (addr handle word) <- get body, data
-  # *curr-word = "x"
-  allocate curr-word-ah
-  var tmp/eax: (addr word) <- lookup *curr-word-ah
-  curr-word <- copy tmp
-  initialize-word-with curr-word, "x"
-  # *curr-word->next = "1"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "1"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  # curr-word = curr-word->next
-  curr-word-ah <- copy next-word-ah
-  curr-word <- copy tmp
-  # *curr-word->next = "+"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "+"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  tmp <- lookup *prev-word-ah
+  curr-word-ah <- get body, data
+  parse-words "x 1 +", curr-word-ah
   # x 2+ = x 1+ 1+
   var next/esi: (addr handle function) <- get f, next
   allocate next
@@ -193,32 +145,8 @@ fn create-primitive-functions _self: (addr handle function) {
   allocate body-ah
   var body/eax: (addr line) <- lookup *body-ah
   initialize-line body
-  var curr-word-ah/ecx: (addr handle word) <- get body, data
-  # *curr-word = "x"
-  allocate curr-word-ah
-  var tmp/eax: (addr word) <- lookup *curr-word-ah
-  curr-word <- copy tmp
-  initialize-word-with curr-word, "x"
-  # *curr-word->next = "1+"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "1+"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  # curr-word = curr-word->next
-  curr-word-ah <- copy next-word-ah
-  curr-word <- copy tmp
-  # *curr-word->next = "1+"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "1+"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  tmp <- lookup *prev-word-ah
+  curr-word-ah <- get body, data
+  parse-words "x 1+ 1+", curr-word-ah
   # x square = x x *
   var next/esi: (addr handle function) <- get f, next
   allocate next
@@ -234,32 +162,8 @@ fn create-primitive-functions _self: (addr handle function) {
   allocate body-ah
   var body/eax: (addr line) <- lookup *body-ah
   initialize-line body
-  var curr-word-ah/ecx: (addr handle word) <- get body, data
-  # *curr-word = "x"
-  allocate curr-word-ah
-  var tmp/eax: (addr word) <- lookup *curr-word-ah
-  var curr-word/edx: (addr word) <- copy tmp
-  initialize-word-with curr-word, "x"
-  # *curr-word->next = "x"
-  var next-word-ah/ebx: (addr handle word) <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "x"
-  # *curr-word->next->prev = curr-word
-  var prev-word-ah/edi: (addr handle word) <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  # curr-word = curr-word->next
-  curr-word-ah <- copy next-word-ah
-  curr-word <- copy tmp
-  # *curr-word->next = "*"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "*"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  tmp <- lookup *prev-word-ah
+  curr-word-ah <- get body, data
+  parse-words "x x *", curr-word-ah
   # x 1- = x 1 -
   var next/esi: (addr handle function) <- get f, next
   allocate next
@@ -275,32 +179,8 @@ fn create-primitive-functions _self: (addr handle function) {
   allocate body-ah
   var body/eax: (addr line) <- lookup *body-ah
   initialize-line body
-  var curr-word-ah/ecx: (addr handle word) <- get body, data
-  # *curr-word = "x"
-  allocate curr-word-ah
-  var tmp/eax: (addr word) <- lookup *curr-word-ah
-  curr-word <- copy tmp
-  initialize-word-with curr-word, "x"
-  # *curr-word->next = "1"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "1"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  # curr-word = curr-word->next
-  curr-word-ah <- copy next-word-ah
-  curr-word <- copy tmp
-  # *curr-word->next = "-"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "-"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  tmp <- lookup *prev-word-ah
+  curr-word-ah <- get body, data
+  parse-words "x 1 -", curr-word-ah
   # x y sub = x y -
   var next/esi: (addr handle function) <- get f, next
   allocate next
@@ -321,32 +201,8 @@ fn create-primitive-functions _self: (addr handle function) {
   allocate body-ah
   var body/eax: (addr line) <- lookup *body-ah
   initialize-line body
-  var curr-word-ah/ecx: (addr handle word) <- get body, data
-  # *curr-word = "x"
-  allocate curr-word-ah
-  var tmp/eax: (addr word) <- lookup *curr-word-ah
-  curr-word <- copy tmp
-  initialize-word-with curr-word, "x"
-  # *curr-word->next = "y"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "y"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  # curr-word = curr-word->next
-  curr-word-ah <- copy next-word-ah
-  curr-word <- copy tmp
-  # *curr-word->next = "-"
-  next-word-ah <- get curr-word, next
-  allocate next-word-ah
-  tmp <- lookup *next-word-ah
-  initialize-word-with tmp, "-"
-  # *curr-word->next->prev = curr-word
-  prev-word-ah <- get tmp, prev
-  copy-object curr-word-ah, prev-word-ah
-  tmp <- lookup *prev-word-ah
+  curr-word-ah <- get body, data
+  parse-words "x y -", curr-word-ah
 }
 
 fn function-body functions: (addr handle function), _word: (addr handle word), out: (addr handle line) {
diff --git a/apps/tile/word.mu b/apps/tile/word.mu
index 69142234..9baa4b94 100644
--- a/apps/tile/word.mu
+++ b/apps/tile/word.mu
@@ -4,7 +4,6 @@ fn initialize-word _self: (addr word) {
   allocate data-ah
   var data/eax: (addr gap-buffer) <- lookup *data-ah
   initialize-gap-buffer data
-  # TODO: sometimes initialize box-data rather than scalar-data
 }
 
 ## some helpers for creating words. mostly for tests
@@ -571,3 +570,31 @@ fn word-list-length words: (addr handle word) -> _/eax: int {
   }
   return result
 }
+
+# out-ah already has a word allocated and initialized
+fn parse-words in: (addr array byte), out-ah: (addr handle word) {
+  var in-stream: (stream byte 0x100)
+  var in-stream-a/esi: (addr stream byte) <- address in-stream
+  write in-stream-a, in
+  var cursor-word-ah/ebx: (addr handle word) <- copy out-ah
+  $parse-words:loop: {
+    var done?/eax: boolean <- stream-empty? in-stream-a
+    compare done?, 0  # false
+    break-if-!=
+    var _g/eax: grapheme <- read-grapheme in-stream-a
+    var g/ecx: grapheme <- copy _g
+    # if not space, insert
+    compare g, 0x20  # space
+    {
+      break-if-=
+      var cursor-word/eax: (addr word) <- lookup *cursor-word-ah
+      add-grapheme-to-word cursor-word, g
+      loop $parse-words:loop
+    }
+    # otherwise insert word after and move cursor to it
+    append-word cursor-word-ah
+    var cursor-word/eax: (addr word) <- lookup *cursor-word-ah
+    cursor-word-ah <- get cursor-word, next
+    loop
+  }
+}