CSV Exporter (#227)

* CSV Exporter

* CSV Exporter

* updated readme

---------

Co-authored-by: Pavel Zhukov <33721692+LeaveMyYard@users.noreply.github.com>

3 files changed, 113 insertions, 0 deletions

diff --git a/README.md b/README.md
index 3a34a1a..908310d 100644
--- a/README.md
+++ b/README.md
@@ -285,6 +285,7 @@ Currently KRR ships with a few formatters to represent the scan data:
 - `json`
 - `yaml`
 - `pprint` - data representation from python's pprint library
+- `csv_export` - export data to a csv file in the current directory
 
 To run a strategy with a selected formatter, add a `-f` flag:
 
diff --git a/robusta_krr/formatters/__init__.py b/robusta_krr/formatters/__init__.py
index 325cf01..e34a25f 100644
--- a/robusta_krr/formatters/__init__.py
+++ b/robusta_krr/formatters/__init__.py
@@ -2,3 +2,4 @@ from .json import json
 from .pprint import pprint
 from .table import table
 from .yaml import yaml
+from .csv import csv
diff --git a/robusta_krr/formatters/csv.py b/robusta_krr/formatters/csv.py
new file mode 100644
index 0000000..792c061
--- /dev/null
+++ b/robusta_krr/formatters/csv.py
@@ -0,0 +1,111 @@
+import itertools
+import csv
+
+import logging
+
+
+from robusta_krr.core.abstract import formatters
+from robusta_krr.core.models.allocations import RecommendationValue
+from robusta_krr.core.models.result import ResourceScan, ResourceType, Result
+from robusta_krr.utils import resource_units
+import datetime
+
+logger = logging.getLogger("krr")
+
+NONE_LITERAL = "unset"
+NAN_LITERAL = "?"
+
+
+def _format(value: RecommendationValue) -> str:
+    if value is None:
+        return NONE_LITERAL
+    elif isinstance(value, str):
+        return NAN_LITERAL
+    else:
+        return resource_units.format(value)
+
+
+def __calc_diff(allocated, recommended, selector, multiplier=1) -> str:
+    if recommended is None or isinstance(recommended.value, str) or selector != "requests":
+        return ""
+    else:
+        reccomended_val = recommended.value if isinstance(recommended.value, (int, float)) else 0
+        allocated_val = allocated if isinstance(allocated, (int, float)) else 0
+        diff_val = reccomended_val - allocated_val
+        diff_sign = "+" if diff_val >= 0 else "-"
+        return f"{diff_sign}{_format(abs(diff_val) * multiplier)}"
+
+
+def _format_request_str(item: ResourceScan, resource: ResourceType, selector: str) -> str:
+    allocated = getattr(item.object.allocations, selector)[resource]
+    recommended = getattr(item.recommended, selector)[resource]
+
+    if allocated is None and recommended.value is None:
+        return f"{NONE_LITERAL}"
+
+    diff = __calc_diff(allocated, recommended, selector)
+    if diff != "":
+        diff = f"({diff}) "
+
+    return (
+        diff
+        + _format(allocated)
+        + " -> "
+        + _format(recommended.value)
+    )
+
+
+def _format_total_diff(item: ResourceScan, resource: ResourceType, pods_current: int) -> str:
+    selector = "requests"
+    allocated = getattr(item.object.allocations, selector)[resource]
+    recommended = getattr(item.recommended, selector)[resource]
+
+    return __calc_diff(allocated, recommended, selector, pods_current)
+
+
+@formatters.register()
+def csv_export(result: Result) -> str:
+    
+    current_datetime = datetime.datetime.now().strftime("%Y%m%d%H%M%S")
+    file_path = f"krr-{current_datetime}.csv"
+
+    # We need to order the resource columns so that they are in the format of Namespace,Name,Pods,Old Pods,Type,Container,CPU Diff,CPU Requests,CPU Limits,Memory Diff,Memory Requests,Memory Limits
+    resource_columns = []
+    for resource in ResourceType:
+        resource_columns.append(f"{resource.name} Diff")
+        resource_columns.append(f"{resource.name} Requests")
+        resource_columns.append(f"{resource.name} Limits")
+
+    with open(file_path, 'w+', newline='') as csvfile:
+        csv_writer = csv.writer(csvfile)
+        csv_writer.writerow([
+            "Namespace", "Name", "Pods", "Old Pods", "Type", "Container",
+            *resource_columns
+
+        ])
+
+        for _, group in itertools.groupby(
+            enumerate(result.scans), key=lambda x: (x[1].object.cluster, x[1].object.namespace, x[1].object.name)
+        ):
+            group_items = list(group)
+
+            for j, (i, item) in enumerate(group_items):
+                full_info_row = j == 0
+
+                row = [
+                    item.object.namespace if full_info_row else "",
+                    item.object.name if full_info_row else "",
+                    f"{item.object.current_pods_count}" if full_info_row else "",
+                    f"{item.object.deleted_pods_count}" if full_info_row else "",
+                    item.object.kind if full_info_row else "",
+                    item.object.container,
+                ]
+
+                for resource in ResourceType:
+                    row.append(_format_total_diff(item, resource, item.object.current_pods_count))
+                    row += [_format_request_str(item, resource, selector) for selector in ["requests", "limits"]]
+
+                csv_writer.writerow(row)
+             
+    logger.info("CSV File: %s", file_path)
+    return ""
\ No newline at end of file