Progress of Direct Read TTF (finished profile information, entering 2D stage, 50%)

The correct curve (progress made at a.m. on November 8,) does not contain the curve. (This progress was made at on May 5)

There is no curve in the font of the simhei. I suddenly thought of, and the generation without curves is faster. I will use next time.

I did a lot of tests, mainly wondering that there is too little information in the description of the TrueType curve format. The comrades who have done so have taken a pen. No one will explain it in detail, what is the explanation of the long string of TrueType that is no longer on the curve. Now, I want to draw a curve in the order of 012 234 456 according to the requirements of second-order besell.

You have to study FreeType'sSource codeTo see if this is the case, at least the current vision is correct.

The usage is as follows. First, obtain a template from the file.

During render2path, A fontcurvelist object is created from the module information.

Then render, and finally retrieve the pathlist, which is the vertex list. Use this information to create vertexbuffer.

The render parameter is used to specify the distance between the beiser curve segments. I use a curve to segment the data according to the length of the Curve. For example, if 0.2 is input, a vertex is created at every 0.2 distance on a curve. Actually uploading 0.2 points is too small, resulting in a large number of vertices.

FreeType has a large space by default. For the specific size, you can see the fontsize member in fontcurvelist. As for the input ''dry '', It is a redundant parameter and you forgot to delete it.

The next step is to design another clfont_render2texture.

History (from 2010.5.2, every day to am)

Error result.

This is mainly because the curve is not drawn separately. In addition, The 0th vertices of TrueType are completely for the mark position and are not used for plotting. It is correct.

However, I cannot output a normal Chinese Character using simsunb. TTF, which is very uncommon.

At first, I thought that my cmap table was incorrectly parsed. I checked it for one night and couldn't find anything wrong. Also found by the way information, this font cmap table uses a OpenType cmap format12 extension, with the USC-4.

Hard For A Long Time handwritten Utf-8 to USC-4 conversion, WP7 SDK UTF32 don't let use, some strange. Later it was found that the cmap table of simsunb. TTF records the text after 0x20000.

A check is Japan and South Korea's uncommon words..., it turned out to be a CJK-B extension font.

A dream of life ....

After changing the font, I tried several small font libraries and some words would fail. Some non-existent words will be placed in the cmap table. I don't know if I should find if it doesn't exist.

Then I found the old-fashioned simhei and simhei, and finally drew them normally.

Correct results

Use pure C # To ensure that Windows Phone 7 can be transplanted. The efficiency must be unsafe.CodeIncomparable.

In addition, it is found that DOTNET framework is used in Windows Phone 7. Microsoft is not tired.

In PC, let me use 4, and in Windows Phone 7, let me use 3. Although it is similar, I still have to compile it and try it. Are you doing this to promote your ability to create a copy project?

Now I have created a copy version of Windows Phone 7 project, which requires code compatibility at any time.

The number of characters read is 30%. If you want to use 3D text, you can use it. However, in 2D, there is still a process to obtain the typographical information of characters.

I don't know if Microsoft will make a font package for the DOTNET version. I feel more and more that I am doing what Microsoft should do.