table fixes and notes

core/src/processing/data/Table.java

@@ -159,6 +159,7 @@ public class Table {
     for (TableRow row : rows) {
       if (!typed) {
         setColumnTypes(row.getColumnTypes());
+        typed = true;
       }
       addRow(row);
     }
@@ -417,6 +418,10 @@ public class Table {
     char[] c = new char[100];
     int count = 0;
     boolean insideQuote = false;
+
+    int alloc = 100;
+    setRowCount(100);
+
     int row = 0;
     int col = 0;
     int ch;
@@ -462,14 +467,23 @@ public class Table {
           }
           setString(row, col, new String(c, 0, count));
           count = 0;
-          if (row == 0 && header) {
+          row++;
+          if (row == 1 && header) {
             // Use internal row removal (efficient because only one row).
             removeTitleRow();
             // Un-set the header variable so that next time around, we don't
             // just get stuck into a loop, removing the 0th row repeatedly.
             header = false;
+            // Reset the number of rows (removeTitleRow() won't reset our local 'row' counter)
+            row = 0;
+          }
+//          if (row % 1000 == 0) {
+//            PApplet.println(PApplet.nfc(row));
+//          }
+          if (row == alloc) {
+            alloc *= 2;
+            setRowCount(alloc);
           }
-          row++;
           col = 0;
 
         } else if (ch == ',') {
@@ -491,6 +505,9 @@ public class Table {
     if (count > 0) {
       setString(row, col, new String(c, 0, count));
     }
+    if (alloc != row) {
+      setRowCount(row);  // shrink to the actual size
+    }
   }
 
 

core/todo.txt

@@ -6,6 +6,12 @@ X   https://github.com/processing/processing/issues/257
 X   PVector discussion with Dan
 o   Jer and Dan will look at their code, plus toxiclibs
 
+table
+X major performance improvements to 'newlines' parsing
+X debugging table parsing with header rows
+X bug fix for setting data type
+
+
 applet removal
 _ remove Applet as base class
 _ performance issues on OS X (might be threading due to Applet)
@@ -73,6 +79,8 @@ _   hint(ENABLE_RETINA_PIXELS) or hint(ENABLE_HIDPI_PIXELS)
 _   hint(ENABLE_2X_PIXELS)?
 _   hidpi is Apple's name as well
 _ no high-res display support for OpenGL
+_   https://github.com/processing/processing/issues/2573
+_   https://jogamp.org/bugzilla/show_bug.cgi?id=741
 _ no high-dpi support for core on Windows
 _   https://github.com/processing/processing/issues/2411
 _ retina sketches slow to start

todo.txt

@@ -3,12 +3,19 @@ X increase heap size to 256m (-Xmx256) per Manindra request
 X use a ButtonGroup so that the current Mode cannot be de-selected
 X   https://github.com/processing/processing/issues/2545
 
-_ color selector for the pref
-_   https://github.com/processing/processing/pull/2568
 
 pulls
 X Implementation of a list of open sketches in the Sketch menu
 X   https://github.com/processing/processing/pull/2551
+_ color selector for the pref (not reading it properly)
+_   https://github.com/processing/processing/pull/2568
+
+_ look at the sound library https://github.com/wirsing/ProcessingSound
+
+_ moving p5 examples to the web repo
+all examples are out of "processing/java" and are now in "processing-docs/content/". The Book examples have been removed entirely from our repositories.
+_ remove reference.zip from main repo
+_ processing-web/java_generate/ReferenceGenerator
 
 _ try new syntax package
 _ possibility of libraries folder inside a particular sketch?
@@ -23,8 +30,6 @@ _ move examples into web repo
 _ make reference and examples build process part of dist
 _   separate ant targets, but only require them for dist
 _   as separate targets, folks can build explicitly if they'd like
-_ remove reference.zip from main repo
-_ processing-web/java_generate/ReferenceGenerator
 
 _ shouldn't write sketch.properties unless it's a non-default mode
 _   https://github.com/processing/processing/issues/2531