From a727e38db47105528ffb944261fb31e1849e8fbc Mon Sep 17 00:00:00 2001
From: Florez Ospina Juan Felipe <juan.florez-ospina@psi.ch>
Date: Tue, 26 Mar 2024 16:14:40 +0100
Subject: [PATCH] Implemented hdf5_vis.py, which is a hdf5 visualization
 library to obtain treemap and yaml representations of hdf5 files.

---
 src/hdf5_vis.py | 93 ++++++++++++++++++++++++++++++++-----------------
 1 file changed, 62 insertions(+), 31 deletions(-)

diff --git a/src/hdf5_vis.py b/src/hdf5_vis.py
index 636f452..ada825a 100644
--- a/src/hdf5_vis.py
+++ b/src/hdf5_vis.py
@@ -1,22 +1,19 @@
+import sys
+import os
+root_dir = os.path.abspath(os.curdir)
+sys.path.append(root_dir)
+
 import h5py
 import yaml
-import os
+
+import src.hdf5_lib as hdf5_lib
+
 import numpy as np
-import numpy as pd
+import pandas as pd
 
-import config_file
-import hdf5_lib
-import g5505_utils as utils
-
-import matplotlib.pyplot as plt
-import plotly.express as px
-import plotly.graph_objects as go
 from plotly.subplots import make_subplots
-
-import subprocess
-
-#output = subprocess.run("git status",capture_output=True)
-#output.stdout()
+import plotly.graph_objects as go
+import plotly.express as px
 
 def display_group_hierarchy_on_a_treemap(filename: str):
 
@@ -63,9 +60,12 @@ def display_group_hierarchy_on_a_treemap(filename: str):
 def make_dtype_yaml_compatible(value):
     try:
         if isinstance(value, np.generic):
-            if np.issubdtype(value.dtype, np.string_) or np.issubdtype(value.dtype, np.unicode_):
+            #if np.issubdtype(value.dtype, np.string_):
                 #value = value.astype(str)
-                value = str(value)
+            if np.issubdtype(value.dtype, np.bytes_):
+                value = value.decode('utf-8')
+            elif np.issubdtype(value.dtype, np.unicode_):
+                value = str(value) 
             elif np.issubdtype(value.dtype, np.number):
                 value = float(value)
             else:
@@ -91,30 +91,49 @@ def make_dtype_yaml_compatible(value):
 
     return value
 
-def print_metadata(name, obj, yaml_dict): 
+def construct_attributes_dict(attrs_obj):
+
+    attr_dict = {}
+    for key, value in attrs_obj.items():
+        if not key in ['file_list','filtered_file_list']:
+
+            value =  make_dtype_yaml_compatible(value)
+
+            #if isinstance(value,str):
+            #    value.replace('\\','\\\\')
+
+            attr_dict[key] = {"rename_as" : key,
+                              "value" : value
+                             }
+    return attr_dict
+
+def print_metadata(name, obj, folder_depth, yaml_dict): 
 
     # TODO: should we enable deeper folders ?
-    if len(obj.name.split('/')) <= 4:
+    if len(obj.name.split('/')) <= folder_depth:
         name_to_list = obj.name.split('/')
         name_head = name_to_list[-1]
 
         if isinstance(obj,h5py.Group):
             #print('name:', obj.name)
             #print('attributes:', dict(obj.attrs))
-            attr_dict = {}
+            #attr_dict = {}
             group_dict = {}
-            for key, value in obj.attrs.items():
+
+            attr_dict = construct_attributes_dict(obj.attrs)
+
+            #for key, value in obj.attrs.items():
                 #print (key, value.dtype)
-                if key == 'Layout':
-                    print(value)
+            #    if key == 'Layout':
+            #        print(value)
 
-                if not key in ['file_list','filtered_file_list']:
+            #    if not key in ['file_list','filtered_file_list']:
 
-                    value =  make_dtype_yaml_compatible(value)
+            #        value =  make_dtype_yaml_compatible(value)
 
-                    attr_dict[key] = {'rename_as' : key,
-                                      'value' : value
-                                      }
+            #        attr_dict[key] = {'rename_as' : key,
+            #                          'value' : value
+            #                          }
             
             #group_dict[obj.name] = {'name': obj.name, 'attributes': attr_dict}
             group_dict = {"name": name_head, "attributes": attr_dict, "datasets":{}}
@@ -126,7 +145,7 @@ def print_metadata(name, obj, yaml_dict):
             yaml_dict[obj.name] = group_dict
         elif isinstance(obj, h5py.Dataset):            
             parent_name = '/'.join(name_to_list[:-1])
-            yaml_dict[parent_name]["datasets"][name_head] = {'rename_as': name_head ,'attributes':dict(obj.attrs)}
+            yaml_dict[parent_name]["datasets"][name_head] = {"rename_as": name_head ,"attributes":dict(obj.attrs)}
             #print(yaml.dump(group_dict,sort_keys=False))
 
         #elif len(obj.name.split('/')) == 3:
@@ -135,18 +154,30 @@ def print_metadata(name, obj, yaml_dict):
 
 
 
-def take_yml_snapshot_of_hdf5_file(input_filename_path):
+def take_yml_snapshot_of_hdf5_file(input_filename_path,folder_depth: int = 4):
 
     yaml_dict = {}
 
     output_filename_tail, ext = os.path.splitext(input_filename_path)
 
     with h5py.File(input_filename_path,'r') as f:
-        f.visititems(lambda name, obj: print_metadata(name,obj,yaml_dict))
+
+        attrs_dict = construct_attributes_dict(f.attrs)
+        yaml_dict[f.name] =  {"name": f.name, "attributes": attrs_dict, "datasets":{}}
+        f.visititems(lambda name, obj: print_metadata(name,obj,folder_depth,yaml_dict))
+
+    #with open(output_filename_tail+".json","w") as yaml_file:
+    #    json_obj = json.dumps(yaml_dict,indent=4,sort_keys=False,)
+    #    yaml_file.write(json_obj)
 
     with open(output_filename_tail+".yaml","w") as yaml_file:
-        yaml.dump(yaml_dict,yaml_file,sort_keys=False)
+        yaml_output = yaml.dump(yaml_dict,sort_keys=False)
+        #for key in yaml_dict:
+        #    yaml_output = yaml.dump(yaml_dict[key],sort_keys=False)
+        yaml_file.write(yaml_output )
 
     return output_filename_tail+".yaml"
 
 
+
+