This is an automated email from the ASF dual-hosted git repository.
morningman pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris-website.git
The following commit(s) were added to refs/heads/master by this push:
new fb967f3e800 [doc] Add type of SET for BitmapValue (#1189)
fb967f3e800 is described below
commit fb967f3e8009577afcc970a4feff7163938ed28a
Author: Jerry Hu <[email protected]>
AuthorDate: Wed Oct 16 23:22:51 2024 +0800
[doc] Add type of SET for BitmapValue (#1189)
# Versions
- [x] dev
- [x] 3.0
- [x] 2.1
- [ ] 2.0
# Languages
- [x] Chinese
- [x] English
---
community/developer-guide/bitmap-hll-file-format.md | 1 +
.../current/developer-guide/bitmap-hll-file-format.md | 1 +
2 files changed, 2 insertions(+)
diff --git a/community/developer-guide/bitmap-hll-file-format.md
b/community/developer-guide/bitmap-hll-file-format.md
index 201e99a7e26..6296d57e9e6 100644
--- a/community/developer-guide/bitmap-hll-file-format.md
+++ b/community/developer-guide/bitmap-hll-file-format.md
@@ -45,6 +45,7 @@ The Flag value description is as follows:
| 2 | BITMAP32, 32-bit bitmap corresponds to the type
`org.roaringbitmap.RoaringBitmap` in java. The type is `roaring::Roaring` in
C++, and the following data is the structure after the sequence of
roaring::Roaring. You can use `org in java. .roaringbitmap.RoaringBitmap` or
`roaring::Roaring` in c++ directly deserialize |
| 3 | SINGLE64, there is only one 64-bit unsigned integer value in the bitmap,
and the next 8 bytes represent the 64-bit unsigned integer value |
| 4 | BITMAP64, 64-bit bitmap corresponds to `org.roaringbitmap.RoaringBitmap`
in java; `Roaring64Map` in doris in c++. The data structure is the same as the
result in the roaring library, but the serialization and deserialization
methods It is different, there will be 1-8 bytes of variable-length encoding
uint64 in the bitmap representation of the size. The following data is a series
of multiple high-order representations of 4 bytes and 32-bit roaring bitmap
serialized data repeated |
+| 5 | SET, When the number of values in the bitmap is greater than one and
less than 32, the BitmapValue actually uses a hash set to store the values. The
data structure: A uint8_t of one byte represents the number of values, followed
by the values themselves (8 bytes each, uint64_t). ｜
C++ serialization and deserialization examples are in the
`BitmapValue::write()` method in `be/src/util/bitmap_value.h` and the Java
examples are in the `serialize()` `deserialize()` method in
`fe/fe-common/src/main/java/org/apache/doris/common/io/BitmapValue.java`.
diff --git
a/i18n/zh-CN/docusaurus-plugin-content-docs-community/current/developer-guide/bitmap-hll-file-format.md
b/i18n/zh-CN/docusaurus-plugin-content-docs-community/current/developer-guide/bitmap-hll-file-format.md
index 8e2aacd0c14..9dac330d7ea 100644
---
a/i18n/zh-CN/docusaurus-plugin-content-docs-community/current/developer-guide/bitmap-hll-file-format.md
+++
b/i18n/zh-CN/docusaurus-plugin-content-docs-community/current/developer-guide/bitmap-hll-file-format.md
@@ -45,6 +45,7 @@ Flag 值说明如下：
| 2 | BITMAP32，32 位bitmap 对应java 中类型为 `org.roaringbitmap.RoaringBitmap` C++
中类型为`roaring::Roaring`， 后面data 为roaring::Roaring 序列后的结构， 可以使用java 中的
`org.roaringbitmap.RoaringBitmap` 或c++中`roaring::Roaring` 直接反序列化 |
| 3 | SINGLE64 ,bitmap 中只有一个64位无符号整型值，后面8个字节表示64位无符号整型值 |
| 4 | BITMAP64, 64 位bitmap 对应java 中类型为 `org.roaringbitmap.RoaringBitmap;`
c++ 中类型为doris 中的`Roaring64Map`，数据结构和 roaring
库中的结果一致，但是序列化和反序列话方法有所不同，后面会有1-8个字节的变长编码的uint64 的表述bitmap 中size。后面的数据是各式是多个
4个字节的高位表示和32位 roaring bitmap 序列化数据重复而成 |
+| 5 | SET, 当 bitmap 的值的个数在1和32之间时, 实际的存储类型是 hashset. 数据结构是: 一个字节代表值的个数,
后面跟着所有值（每个值的长度是8字节）。 |
c++ 序列化和反序列化的示例 在 `be/src/util/bitmap_value.h` 的`BitmapValue::write()`
`BitmapValue::deserialize()` 方法中, Java示例在
`fe/fe-common/src/main/java/org/apache/doris/common/io/BitmapValue.java`
中的`serialize()` `deserialize()` 方法中。
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
