sub_F01_par_load_count.f90

中文 English

直接用途

sub_F01_par_load_count 读取指定粒子文件的列数/粒子数，用于调用者分配或检查粒子数组容量。

参数表

参数	方向	shape/范围	含义与局部约定
tag	in	scalar or caller-provided array	I/O 格式选择标签，调度到 dat/bin/h5 等具体读写例程。
label	in	scalar or caller-provided array	文件目录名和文件名前缀；通常是简单名字，不应包含路径分隔符。
it	in	scalar or caller-provided array	时间步或迭代编号，会按固定宽度编码到文件名中。
nvar	in	scalar or caller-provided array	粒子变量数或文件中的变量列数。
np	out	scalar or caller-provided array	粒子数；读写例程只处理 par(:,1:np)，碰撞/交换例程可能更新它。

局部假设

I/O 例程不改变数据物理含义，只按约定文件名和数组内存顺序读写。raw .bin 文件使用默认 Fortran real 的 stream 数据；读写双方必须使用一致的实数精度、端序和数组 shape。HDF5 入口依赖 USE_HDF5 构建宏和 HDF5 模块。

实现逻辑

调度入口根据 tag 选择具体文件格式例程，不直接改变数组内容。

调用注意

本页只说明该例程本身的调用边界和实现事实；完整接口声明和源码级 API 仍以英文侧的 generated API 为准。

Direct Purpose

sub_F01_par_load_count get local particle count np from per-rank dat/bin/h5 file.

Parameter Table

Parameter	Direction	Shape/range	Meaning and local convention
tag	in	scalar or caller-provided array	‘dat’|’bin’|’h5’|’hdf5’
label	in	scalar or caller-provided array	directory and file prefix
it	in	scalar or caller-provided array	iteration index encoded in file name
nvar	in	scalar or caller-provided array	number of particle variables (required for ‘bin’; optional check for ‘h5’; ignored for ‘dat’)
np	out	scalar or caller-provided array	particle count inferred from file

Local Assumptions

I/O routines do not change the physical meaning of data; they only read or write arrays using the naming and memory-order conventions. Raw .bin files store default Fortran real stream data, so readers and writers must agree on real kind, endianness, and array shape. HDF5 entries depend on the USE_HDF5 build macro and HDF5 module.

Implementation Notes

The dispatch entry selects the concrete file-format routine from tag and does not otherwise transform the array content.

Generated API

Get local particle count np from per-rank dat/bin/h5 file.

File name pattern: label/label_time_rank.EXT time = it formatted as I10.10, rank = MPI rank formatted as I5.5.

Counting method:

• tag=’dat’ : count non-empty lines.

• tag=’bin’ : np = filesize_bytes / (nvar * bytes_per_default_real).

• tag=’h5’ : np = second dimension of dataset “par”.

• tag=’hdf5’ : same as ‘h5’.

Author

Yinjian ZHAO (2026/03/03).

param tag:

[in] : ‘dat’|’bin’|’h5’|’hdf5’

param label:

[in] : directory and file prefix

param it:

[in] : iteration index encoded in file name

param nvar:

[in] : number of particle variables (required for ‘bin’; optional check for ‘h5’; ignored for ‘dat’)

param np:

[out] : particle count inferred from file

Functions

subroutine sub_f01_load_par_count(tag, label, it, nvar, np)

namespace hdf5